Datasets:

COLING

/

COLING2024-papers

like 0

Follow

International Conference on Computational Linguistics 1

https://aclanthology.org/2024.lrec-main.401.bib

https://aclanthology.org/2024.lrec-main.401/

@inproceedings{sun-etal-2024-decoding, title = "Decoding at the Speed of Thought: Harnessing Parallel Decoding of Lexical Units for {LLM}s", author = "Sun, Chenxi and Zhang, Hongzhi and Lin, Zijia and Zhang, Jingyuan and Zhang, Fuzheng and Wang, Zhongyuan and Chen, Bin and Song, Chengru and Zhang, Di and Gai, Kun and Xiong, Deyi", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.401", pages = "4476--4487", abstract = "Large language models have demonstrated exceptional capability in natural language understanding and generation. However, their generation speed is limited by the inherently sequential nature of their decoding process, posing challenges for real-time applications. This paper introduces Lexical Unit Decoding (LUD), a novel decoding methodology implemented in a data-driven manner, accelerating the decoding process without sacrificing output quality. The core of our approach is the observation that a pre-trained language model can confidently predict multiple contiguous tokens, forming the basis for a lexical unit, in which these contiguous tokens could be decoded in parallel. Extensive experiments validate that our method substantially reduces decoding time while maintaining generation quality, i.e., 33{\%} speed up on natural language generation with no quality loss, and 30{\%} speed up on code generation with a negligible quality loss of 3{\%}. Distinctively, LUD requires no auxiliary models and does not require changes to existing architectures. It can also be integrated with other decoding acceleration methods, thus achieving an even more pronounced inference efficiency boost. We posit that the foundational principles of LUD could define a new decoding paradigm for future language models, enhancing their applicability for a broader spectrum of applications. All codes are be publicly available at https://github.com/tjunlp-lab/Lexical-Unit-Decoding-LUD-.", }

Large language models have demonstrated exceptional capability in natural language understanding and generation. However, their generation speed is limited by the inherently sequential nature of their decoding process, posing challenges for real-time applications. This paper introduces Lexical Unit Decoding (LUD), a novel decoding methodology implemented in a data-driven manner, accelerating the decoding process without sacrificing output quality. The core of our approach is the observation that a pre-trained language model can confidently predict multiple contiguous tokens, forming the basis for a lexical unit, in which these contiguous tokens could be decoded in parallel. Extensive experiments validate that our method substantially reduces decoding time while maintaining generation quality, i.e., 33{\%} speed up on natural language generation with no quality loss, and 30{\%} speed up on code generation with a negligible quality loss of 3{\%}. Distinctively, LUD requires no auxiliary models and does not require changes to existing architectures. It can also be integrated with other decoding acceleration methods, thus achieving an even more pronounced inference efficiency boost. We posit that the foundational principles of LUD could define a new decoding paradigm for future language models, enhancing their applicability for a broader spectrum of applications. All codes are be publicly available at https://github.com/tjunlp-lab/Lexical-Unit-Decoding-LUD-.

Decoding at the Speed of Thought: Harnessing Parallel Decoding of Lexical Units for LLMs

lrec-main.401

Poster

2405.15208

https://huggingface.co/papers/2405.15208

https://aclanthology.org/2024.lrec-main.402.bib

https://aclanthology.org/2024.lrec-main.402/

@inproceedings{he-etal-2024-decoding, title = "Decoding Probing: Revealing Internal Linguistic Structures in Neural Language Models Using Minimal Pairs", author = "He, Linyang and Chen, Peili and Nie, Ercong and Li, Yuanning and Brennan, Jonathan R.", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.402", pages = "4488--4497", abstract = "Inspired by cognitive neuroscience studies, we introduce a novel {``}decoding probing{''} method that uses minimal pairs benchmark (BLiMP) to probe internal linguistic characteristics in neural language models layer by layer. By treating the language model as the brain and its representations as {``}neural activations{''}, we decode grammaticality labels of minimal pairs from the intermediate layers{'} representations. This approach reveals: 1) Self-supervised language models capture abstract linguistic structures in intermediate layers that GloVe and RNN language models cannot learn. 2) Information about syntactic grammaticality is robustly captured through the first third layers of GPT-2 and also distributed in later layers. As sentence complexity increases, more layers are required for learning grammatical capabilities. 3) Morphological and semantics/syntax interface-related features are harder to capture than syntax. 4) For Transformer-based models, both embeddings and attentions capture grammatical features but show distinct patterns. Different attention heads exhibit similar tendencies toward various linguistic phenomena, but with varied contributions.", }

Inspired by cognitive neuroscience studies, we introduce a novel {``}decoding probing{''} method that uses minimal pairs benchmark (BLiMP) to probe internal linguistic characteristics in neural language models layer by layer. By treating the language model as the brain and its representations as {``}neural activations{''}, we decode grammaticality labels of minimal pairs from the intermediate layers{'} representations. This approach reveals: 1) Self-supervised language models capture abstract linguistic structures in intermediate layers that GloVe and RNN language models cannot learn. 2) Information about syntactic grammaticality is robustly captured through the first third layers of GPT-2 and also distributed in later layers. As sentence complexity increases, more layers are required for learning grammatical capabilities. 3) Morphological and semantics/syntax interface-related features are harder to capture than syntax. 4) For Transformer-based models, both embeddings and attentions capture grammatical features but show distinct patterns. Different attention heads exhibit similar tendencies toward various linguistic phenomena, but with varied contributions.

Decoding Probing: Revealing Internal Linguistic Structures in Neural Language Models Using Minimal Pairs

lrec-main.402

Poster

2403.17299

https://aclanthology.org/2024.lrec-main.403.bib

https://aclanthology.org/2024.lrec-main.403/

@inproceedings{zheng-etal-2024-decompose, title = "Decompose, Prioritize, and Eliminate: Dynamically Integrating Diverse Representations for Multimodal Named Entity Recognition", author = "Zheng, Zihao and Zhang, Zihan and Wang, Zexin and Fu, Ruiji and Liu, Ming and Wang, Zhongyuan and Qin, Bing", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.403", pages = "4498--4508", abstract = "Multi-modal Named Entity Recognition, a fundamental task for multi-modal knowledge graph construction, requires integrating multi-modal information to extract named entities from text. Previous research has explored the integration of multi-modal representations at different granularities. However, they struggle to integrate all these multi-modal representations to provide rich contextual information to improve multi-modal named entity recognition. In this paper, we propose DPE-MNER, which is an iterative reasoning framework that dynamically incorporates all the diverse multi-modal representations following the strategy of {``}decompose, prioritize, and eliminate{''}. Within the framework, the fusion of diverse multi-modal representations is \textbf{decomposed} into hierarchically connected fusion layers that are easier to handle. The incorporation of multi-modal information \textbf{prioritizes} transitioning from {``}easy-to-hard{''} and {``}coarse-to-fine{''}. The explicit modeling of cross-modal relevance \textbf{eliminate} the irrelevances that will mislead the MNER prediction. Extensive experiments on two public datasets have demonstrated the effectiveness of our approach.", }

Multi-modal Named Entity Recognition, a fundamental task for multi-modal knowledge graph construction, requires integrating multi-modal information to extract named entities from text. Previous research has explored the integration of multi-modal representations at different granularities. However, they struggle to integrate all these multi-modal representations to provide rich contextual information to improve multi-modal named entity recognition. In this paper, we propose DPE-MNER, which is an iterative reasoning framework that dynamically incorporates all the diverse multi-modal representations following the strategy of {``}decompose, prioritize, and eliminate{''}. Within the framework, the fusion of diverse multi-modal representations is \textbf{decomposed} into hierarchically connected fusion layers that are easier to handle. The incorporation of multi-modal information \textbf{prioritizes} transitioning from {``}easy-to-hard{''} and {``}coarse-to-fine{''}. The explicit modeling of cross-modal relevance \textbf{eliminate} the irrelevances that will mislead the MNER prediction. Extensive experiments on two public datasets have demonstrated the effectiveness of our approach.

Decompose, Prioritize, and Eliminate: Dynamically Integrating Diverse Representations for Multimodal Named Entity Recognition

lrec-main.403

Poster

https://aclanthology.org/2024.lrec-main.404.bib

https://aclanthology.org/2024.lrec-main.404/

@inproceedings{shivagunde-etal-2024-deconstructing, title = "Deconstructing In-Context Learning: Understanding Prompts via Corruption", author = "Shivagunde, Namrata and Lialin, Vladislav and Muckatira, Sherin and Rumshisky, Anna", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.404", pages = "4509--4529", abstract = "The ability of large language models (LLMs) to {``}learn in context{''} based on the provided prompt has led to an explosive growth in their use, culminating in the proliferation of AI assistants such as ChatGPT, Claude, and Bard. These AI assistants are known to be robust to minor prompt modifications, mostly due to alignment techniques that use human feedback. In contrast, the underlying pre-trained LLMs they use as a backbone are known to be brittle in this respect. Building high-quality backbone models remains a core challenge, and a common approach to assessing their quality is to conduct few-shot evaluation. Such evaluation is notorious for being highly sensitive to minor prompt modifications, as well as the choice of specific in-context examples. Prior work has examined how modifying different elements of the prompt can affect model performance. However, these earlier studies tended to concentrate on a limited number of specific prompt attributes and often produced contradictory results. Additionally, previous research either focused on models with fewer than 15 billion parameters or exclusively examined black-box models like GPT-3 or PaLM, making replication challenging. In the present study, we decompose the entire prompt into four components: task description, demonstration inputs, labels, and inline instructions provided for each demonstration. We investigate the effects of structural and semantic corruptions of these elements on model performance. We study models ranging from 1.5B to 70B in size, using ten datasets covering classification and generation tasks. We find that repeating text within the prompt boosts model performance, and bigger models ({\mbox{$\geq$}}30B) are more sensitive to the semantics of the prompt. Finally, we observe that adding task and inline instructions to the demonstrations enhances model performance even when the instructions are semantically corrupted. The code is available at this URL.", }

The ability of large language models (LLMs) to {``}learn in context{''} based on the provided prompt has led to an explosive growth in their use, culminating in the proliferation of AI assistants such as ChatGPT, Claude, and Bard. These AI assistants are known to be robust to minor prompt modifications, mostly due to alignment techniques that use human feedback. In contrast, the underlying pre-trained LLMs they use as a backbone are known to be brittle in this respect. Building high-quality backbone models remains a core challenge, and a common approach to assessing their quality is to conduct few-shot evaluation. Such evaluation is notorious for being highly sensitive to minor prompt modifications, as well as the choice of specific in-context examples. Prior work has examined how modifying different elements of the prompt can affect model performance. However, these earlier studies tended to concentrate on a limited number of specific prompt attributes and often produced contradictory results. Additionally, previous research either focused on models with fewer than 15 billion parameters or exclusively examined black-box models like GPT-3 or PaLM, making replication challenging. In the present study, we decompose the entire prompt into four components: task description, demonstration inputs, labels, and inline instructions provided for each demonstration. We investigate the effects of structural and semantic corruptions of these elements on model performance. We study models ranging from 1.5B to 70B in size, using ten datasets covering classification and generation tasks. We find that repeating text within the prompt boosts model performance, and bigger models ({\mbox{$\geq$}}30B) are more sensitive to the semantics of the prompt. Finally, we observe that adding task and inline instructions to the demonstrations enhances model performance even when the instructions are semantically corrupted. The code is available at this URL.

Deconstructing In-Context Learning: Understanding Prompts via Corruption

lrec-main.404

Poster

2404.02054

https://aclanthology.org/2024.lrec-main.405.bib

https://aclanthology.org/2024.lrec-main.405/

@inproceedings{wang-etal-2024-deem, title = "{DEEM}: Dynamic Experienced Expert Modeling for Stance Detection", author = "Wang, Xiaolong and Wang, Yile and Cheng, Sijie and Li, Peng and Liu, Yang", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.405", pages = "4530--4541", abstract = "Recent work has made a preliminary attempt to use large language models (LLMs) to solve the stance detection task, showing promising results. However, considering that stance detection usually requires detailed background knowledge, the vanilla reasoning method may neglect the domain knowledge to make a professional and accurate analysis. Thus, there is still room for improvement of LLMs reasoning, especially in leveraging the generation capability of LLMs to simulate specific experts (i.e., multi-agents) to detect the stance. In this paper, different from existing multi-agent works that require detailed descriptions and use fixed experts, we propose a Dynamic Experienced Expert Modeling (DEEM) method which can leverage the generated experienced experts and let LLMs reason in a semi-parametric way, making the experts more generalizable and reliable. Experimental results demonstrate that DEEM consistently achieves the best results on three standard benchmarks, outperforms methods with self-consistency reasoning, and reduces the bias of LLMs.", }

Recent work has made a preliminary attempt to use large language models (LLMs) to solve the stance detection task, showing promising results. However, considering that stance detection usually requires detailed background knowledge, the vanilla reasoning method may neglect the domain knowledge to make a professional and accurate analysis. Thus, there is still room for improvement of LLMs reasoning, especially in leveraging the generation capability of LLMs to simulate specific experts (i.e., multi-agents) to detect the stance. In this paper, different from existing multi-agent works that require detailed descriptions and use fixed experts, we propose a Dynamic Experienced Expert Modeling (DEEM) method which can leverage the generated experienced experts and let LLMs reason in a semi-parametric way, making the experts more generalizable and reliable. Experimental results demonstrate that DEEM consistently achieves the best results on three standard benchmarks, outperforms methods with self-consistency reasoning, and reduces the bias of LLMs.

DEEM: Dynamic Experienced Expert Modeling for Stance Detection

lrec-main.405

Poster

2402.15264

https://aclanthology.org/2024.lrec-main.406.bib

https://aclanthology.org/2024.lrec-main.406/

@inproceedings{agarwal-etal-2024-deep, title = "Deep Learning Based Named Entity Recognition Models for Recipes", author = "Agarwal, Ayush and Kapuriya, Janak and Agrawal, Shubham and Konam, Akhil Vamshi and Goel, Mansi and Gupta, Rishabh and Rastogi, Shrey and Niharika, Niharika and Bagler, Ganesh", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.406", pages = "4542--4554", abstract = "Food touches our lives through various endeavors, including flavor, nourishment, health, and sustainability. Recipes are cultural capsules transmitted across generations via unstructured text. Automated protocols for recognizing named entities, the building blocks of recipe text, are of immense value for various applications ranging from information extraction to novel recipe generation. Named entity recognition is a technique for extracting information from unstructured or semi-structured data with known labels. Starting with manually-annotated data of 6,611 ingredient phrases, we created an augmented dataset of 26,445 phrases cumulatively. Simultaneously, we systematically cleaned and analyzed ingredient phrases from RecipeDB, the gold-standard recipe data repository, and annotated them using the Stanford NER. Based on the analysis, we sampled a subset of 88,526 phrases using a clustering-based approach while preserving the diversity to create the machine-annotated dataset. A thorough investigation of NER approaches on these three datasets involving statistical, fine-tuning of deep learning-based language models and few-shot prompting on large language models (LLMs) provides deep insights. We conclude that few-shot prompting on LLMs has abysmal performance, whereas the fine-tuned spaCy-transformer emerges as the best model with macro-F1 scores of 95.9{\%}, 96.04{\%}, and 95.71{\%} for the manually-annotated, augmented, and machine-annotated datasets, respectively.", }

Food touches our lives through various endeavors, including flavor, nourishment, health, and sustainability. Recipes are cultural capsules transmitted across generations via unstructured text. Automated protocols for recognizing named entities, the building blocks of recipe text, are of immense value for various applications ranging from information extraction to novel recipe generation. Named entity recognition is a technique for extracting information from unstructured or semi-structured data with known labels. Starting with manually-annotated data of 6,611 ingredient phrases, we created an augmented dataset of 26,445 phrases cumulatively. Simultaneously, we systematically cleaned and analyzed ingredient phrases from RecipeDB, the gold-standard recipe data repository, and annotated them using the Stanford NER. Based on the analysis, we sampled a subset of 88,526 phrases using a clustering-based approach while preserving the diversity to create the machine-annotated dataset. A thorough investigation of NER approaches on these three datasets involving statistical, fine-tuning of deep learning-based language models and few-shot prompting on large language models (LLMs) provides deep insights. We conclude that few-shot prompting on LLMs has abysmal performance, whereas the fine-tuned spaCy-transformer emerges as the best model with macro-F1 scores of 95.9{\%}, 96.04{\%}, and 95.71{\%} for the manually-annotated, augmented, and machine-annotated datasets, respectively.

Deep Learning Based Named Entity Recognition Models for Recipes

lrec-main.406

Poster

2402.17447

https://aclanthology.org/2024.lrec-main.407.bib

https://aclanthology.org/2024.lrec-main.407/

@inproceedings{xu-etal-2024-deep, title = "Deep Reinforcement Learning-based Dialogue Policy with Graph Convolutional {Q}-network", author = "Xu, Kai and Wang, Zhengyu and Long, Yuxuan and Zhao, Qiaona", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.407", pages = "4555--4565", abstract = "Deep Reinforcement learning (DRL) has been successfully applied to the dialogue policy of task-oriented dialogue systems. However, one challenge in the existing DRL-based dialogue policy methods is their unstructured state-action representations without the ability to learn the relationship between dialogue states and actions. To alleviate this problem, we propose a graph-structured dialogue policy framework for task-oriented dialogue systems. More specifically, we use an unsupervised approach to construct two different bipartite graphs. Then, we generate the user-related and knowledge-related subgraphs based on the matching dialogue sub-states with bipartite graph nodes. A variant of graph convolutional network is employed to encode dialogue subgraphs. After that, we use a bidirectional gated cycle unit (BGRU) and self-attention mechanism to obtain the high-level historical state representations and employ a neural network for the high-level current state representations. The two state representations are joined to learn the action value of dialogue policy. Experiments implemented with different DRL algorithms demonstrate that the proposed framework significantly improves the effectiveness and stability of dialogue policies.", }

Deep Reinforcement learning (DRL) has been successfully applied to the dialogue policy of task-oriented dialogue systems. However, one challenge in the existing DRL-based dialogue policy methods is their unstructured state-action representations without the ability to learn the relationship between dialogue states and actions. To alleviate this problem, we propose a graph-structured dialogue policy framework for task-oriented dialogue systems. More specifically, we use an unsupervised approach to construct two different bipartite graphs. Then, we generate the user-related and knowledge-related subgraphs based on the matching dialogue sub-states with bipartite graph nodes. A variant of graph convolutional network is employed to encode dialogue subgraphs. After that, we use a bidirectional gated cycle unit (BGRU) and self-attention mechanism to obtain the high-level historical state representations and employ a neural network for the high-level current state representations. The two state representations are joined to learn the action value of dialogue policy. Experiments implemented with different DRL algorithms demonstrate that the proposed framework significantly improves the effectiveness and stability of dialogue policies.

Deep Reinforcement Learning-based Dialogue Policy with Graph Convolutional Q-network

lrec-main.407

Poster

https://aclanthology.org/2024.lrec-main.408.bib

https://aclanthology.org/2024.lrec-main.408/

@inproceedings{cho-etal-2024-deep, title = "Deep Reinforcement Learning with Hierarchical Action Exploration for Dialogue Generation", author = "Cho, Itsugun and Takahashi, Ryota and Yanase, Yusaku and Saito, Hiroaki", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.408", pages = "4566--4579", abstract = "Traditionally, approximate dynamic programming is employed in dialogue generation with greedy policy improvement through action sampling, as the natural language action space is vast. However, this practice is inefficient for reinforcement learning (RL) due to the sparsity of eligible responses with high action values, which leads to weak improvement sustained by random sampling. This paper presents theoretical analysis and experiments that reveal the performance of the dialogue policy is positively correlated with the sampling size. To overcome this limitation, we introduce a novel dual-granularity Q-function that explores the most promising response category to intervene in the sampling process. Our approach extracts actions based on a grained hierarchy, thereby achieving the optimum with fewer policy iterations. Additionally, we use offline RL and learn from multiple reward functions designed to capture emotional nuances in human interactions. Empirical studies demonstrate that our algorithm outperforms baselines across automatic metrics and human evaluations. Further testing reveals that our algorithm exhibits both explainability and controllability, as well as generates responses with higher expected rewards.", }

Traditionally, approximate dynamic programming is employed in dialogue generation with greedy policy improvement through action sampling, as the natural language action space is vast. However, this practice is inefficient for reinforcement learning (RL) due to the sparsity of eligible responses with high action values, which leads to weak improvement sustained by random sampling. This paper presents theoretical analysis and experiments that reveal the performance of the dialogue policy is positively correlated with the sampling size. To overcome this limitation, we introduce a novel dual-granularity Q-function that explores the most promising response category to intervene in the sampling process. Our approach extracts actions based on a grained hierarchy, thereby achieving the optimum with fewer policy iterations. Additionally, we use offline RL and learn from multiple reward functions designed to capture emotional nuances in human interactions. Empirical studies demonstrate that our algorithm outperforms baselines across automatic metrics and human evaluations. Further testing reveals that our algorithm exhibits both explainability and controllability, as well as generates responses with higher expected rewards.

Deep Reinforcement Learning with Hierarchical Action Exploration for Dialogue Generation

lrec-main.408

Poster

https://aclanthology.org/2024.lrec-main.409.bib

https://aclanthology.org/2024.lrec-main.409/

@inproceedings{ashraf-etal-2024-defakts, title = "{D}e{F}akt{S}: A {G}erman Dataset for Fine-Grained Disinformation Detection through Social Media Framing", author = "Ashraf, Shaina and Bezzaoui, Isabel and Andone, Ionut and Markowetz, Alexander and Fegert, Jonas and Flek, Lucie", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.409", pages = "4580--4591", abstract = "In today{'}s rapidly evolving digital age, disinformation poses a significant threat to public sentiment and socio-political dynamics. To address this, we introduce a new dataset {``}DeFaktS{''}, designed to understand and counter disinformation within German media. Distinctively curated across various news topics, DeFaktS offers an unparalleled insight into the diverse facets of disinformation. Our dataset, containing 105,855 posts with 20,008 meticulously labeled tweets, serves as a rich platform for in-depth exploration of disinformation{'}s diverse characteristics. A key attribute that sets DeFaktS apart is, its fine-grain annotations based on polarized categories. Our annotation framework, grounded in the textual characteristics of news content, eliminates the need for external knowledge sources. Unlike most existing corpora that typically assign a singular global veracity value to news, our methodology seeks to annotate every structural component and semantic element of a news piece, ensuring a comprehensive and detailed understanding. In our experiments, we employed a mix of classical machine learning and advanced transformer-based models. The results underscored the potential of DeFaktS, with transformer models, especially the German variant of BERT, exhibiting pronounced effectiveness in both binary and fine-grained classifications.", }

In today{'}s rapidly evolving digital age, disinformation poses a significant threat to public sentiment and socio-political dynamics. To address this, we introduce a new dataset {``}DeFaktS{''}, designed to understand and counter disinformation within German media. Distinctively curated across various news topics, DeFaktS offers an unparalleled insight into the diverse facets of disinformation. Our dataset, containing 105,855 posts with 20,008 meticulously labeled tweets, serves as a rich platform for in-depth exploration of disinformation{'}s diverse characteristics. A key attribute that sets DeFaktS apart is, its fine-grain annotations based on polarized categories. Our annotation framework, grounded in the textual characteristics of news content, eliminates the need for external knowledge sources. Unlike most existing corpora that typically assign a singular global veracity value to news, our methodology seeks to annotate every structural component and semantic element of a news piece, ensuring a comprehensive and detailed understanding. In our experiments, we employed a mix of classical machine learning and advanced transformer-based models. The results underscored the potential of DeFaktS, with transformer models, especially the German variant of BERT, exhibiting pronounced effectiveness in both binary and fine-grained classifications.

DeFaktS: A German Dataset for Fine-Grained Disinformation Detection through Social Media Framing

lrec-main.409

Poster

https://aclanthology.org/2024.lrec-main.410.bib

https://aclanthology.org/2024.lrec-main.410/

@inproceedings{ren-etal-2024-deie, title = "{DEIE}: Benchmarking Document-level Event Information Extraction with a Large-scale {C}hinese News Dataset", author = "Ren, Yubing and Cao, Yanan and Li, Hao and Li, Yingjie and Ma, Zixuan ZM and Fang, Fang and Guo, Ping and Ma, Wei", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.410", pages = "4592--4604", abstract = "A text corpus centered on events is foundational to research concerning the detection, representation, reasoning, and harnessing of online events. The majority of current event-based datasets mainly target sentence-level tasks, thus to advance event-related research spanning from sentence to document level, this paper introduces DEIE, a unified large-scale document-level event information extraction dataset with over 56,000+ events and 242,000+ arguments. Three key features stand out: large-scale manual annotation (20,000 documents), comprehensive unified annotation (encompassing event trigger/argument, summary, and relation at once), and emergency events annotation (covering 19 emergency types). Notably, our experiments reveal that current event-related models struggle with DEIE, signaling a pressing need for more advanced event-related research in the future.", }

A text corpus centered on events is foundational to research concerning the detection, representation, reasoning, and harnessing of online events. The majority of current event-based datasets mainly target sentence-level tasks, thus to advance event-related research spanning from sentence to document level, this paper introduces DEIE, a unified large-scale document-level event information extraction dataset with over 56,000+ events and 242,000+ arguments. Three key features stand out: large-scale manual annotation (20,000 documents), comprehensive unified annotation (encompassing event trigger/argument, summary, and relation at once), and emergency events annotation (covering 19 emergency types). Notably, our experiments reveal that current event-related models struggle with DEIE, signaling a pressing need for more advanced event-related research in the future.

DEIE: Benchmarking Document-level Event Information Extraction with a Large-scale Chinese News Dataset

lrec-main.410

Poster

https://aclanthology.org/2024.lrec-main.411.bib

https://aclanthology.org/2024.lrec-main.411/

@inproceedings{du-etal-2024-delan, title = "{DELAN}: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning", author = "Du, Mengfei and Wu, Binhao and Zhang, Jiwen and Fan, Zhihao and Li, Zejun and Luo, Ruipu and Huang, Xuanjing and Wei, Zhongyu", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.411", pages = "4605--4616", abstract = "Vision-and-Language navigation (VLN) requires an agent to navigate in unseen environment by following natural language instruction. For task completion, the agent needs to align and integrate various navigation modalities, including instruction, observation and navigation history. Existing works primarily concentrate on cross-modal attention at the fusion stage to achieve this objective. Nevertheless, modality features generated by disparate uni-encoders reside in their own spaces, leading to a decline in the quality of cross-modal fusion and decision. To address this problem, we propose a Dual-levEL AligNment (DELAN) framework by cross-modal contrastive learning. This framework is designed to align various navigation-related modalities before fusion, thereby enhancing cross-modal interaction and action decision-making. Specifically, we divide the pre-fusion alignment into dual levels: instruction-history level and landmark-observation level according to their semantic correlations. We also reconstruct a dual-level instruction for adaptation to the dual-level alignment. As the training signals for pre-fusion alignment are extremely limited, self-supervised contrastive learning strategies are employed to enforce the matching between different modalities. Our approach seamlessly integrates with the majority of existing models, resulting in improved navigation performance on various VLN benchmarks, including R2R, R4R, RxR and CVDN.", }

Vision-and-Language navigation (VLN) requires an agent to navigate in unseen environment by following natural language instruction. For task completion, the agent needs to align and integrate various navigation modalities, including instruction, observation and navigation history. Existing works primarily concentrate on cross-modal attention at the fusion stage to achieve this objective. Nevertheless, modality features generated by disparate uni-encoders reside in their own spaces, leading to a decline in the quality of cross-modal fusion and decision. To address this problem, we propose a Dual-levEL AligNment (DELAN) framework by cross-modal contrastive learning. This framework is designed to align various navigation-related modalities before fusion, thereby enhancing cross-modal interaction and action decision-making. Specifically, we divide the pre-fusion alignment into dual levels: instruction-history level and landmark-observation level according to their semantic correlations. We also reconstruct a dual-level instruction for adaptation to the dual-level alignment. As the training signals for pre-fusion alignment are extremely limited, self-supervised contrastive learning strategies are employed to enforce the matching between different modalities. Our approach seamlessly integrates with the majority of existing models, resulting in improved navigation performance on various VLN benchmarks, including R2R, R4R, RxR and CVDN.

DELAN: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning

lrec-main.411

Poster

2404.01994

https://aclanthology.org/2024.lrec-main.412.bib

https://aclanthology.org/2024.lrec-main.412/

@inproceedings{he-etal-2024-demonstration, title = "Demonstration Retrieval-Augmented Generative Event Argument Extraction", author = "He, Shiming and Hong, Yu and Yang, Shuai and Yao, Jianmin and Zhou, Guodong", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.412", pages = "4617--4625", abstract = "We tackle Event Argument Extraction (EAE) in the manner of template-based generation. Based on our exploration of generative EAE, it suffers from several issues, such as multiple arguments of one role, generating words out of context and inconsistency with prescribed format. We attribute it to the weakness of following complex input prompts. To address these problems, we propose the demonstration retrieval-augmented generative EAE (DRAGEAE), containing two components: event knowledge-injected generator (EKG) and demonstration retriever (DR). EKG employs event knowledge prompts to capture role dependencies and semantics. DR aims to search informative demonstrations from training data, facilitating the conditional generation of EKG. To train DR, we use the probability-based rankings from large language models (LLMs) as supervised signals. Experimental results on ACE-2005, RAMS and WIKIEVENTS demonstrate that our method outperforms all strong baselines and it can be generalized to various datasets. Further analysis is conducted to discuss the impact of diverse LLMs and prove that our model alleviates the above issues.", }

We tackle Event Argument Extraction (EAE) in the manner of template-based generation. Based on our exploration of generative EAE, it suffers from several issues, such as multiple arguments of one role, generating words out of context and inconsistency with prescribed format. We attribute it to the weakness of following complex input prompts. To address these problems, we propose the demonstration retrieval-augmented generative EAE (DRAGEAE), containing two components: event knowledge-injected generator (EKG) and demonstration retriever (DR). EKG employs event knowledge prompts to capture role dependencies and semantics. DR aims to search informative demonstrations from training data, facilitating the conditional generation of EKG. To train DR, we use the probability-based rankings from large language models (LLMs) as supervised signals. Experimental results on ACE-2005, RAMS and WIKIEVENTS demonstrate that our method outperforms all strong baselines and it can be generalized to various datasets. Further analysis is conducted to discuss the impact of diverse LLMs and prove that our model alleviates the above issues.

Demonstration Retrieval-Augmented Generative Event Argument Extraction

lrec-main.412

Poster

https://aclanthology.org/2024.lrec-main.413.bib

https://aclanthology.org/2024.lrec-main.413/

@inproceedings{pelicon-etal-2024-denoising, title = "Denoising Labeled Data for Comment Moderation Using Active Learning", author = "Pelicon, Andra{\v{z}} and Karan, Mladen and Shekhar, Ravi and Purver, Matthew and Pollak, Senja", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.413", pages = "4626--4633", abstract = "Noisily labeled textual data is ample on internet platforms that allow user-created content. Training models, such as offensive language detection models for comment moderation, on such data may prove difficult as the noise in the labels prevents the model to converge. In this work, we propose to use active learning methods for the purposes of denoising training data for model training. The goal is to sample examples the most informative examples with noisy labels with active learning and send them to the oracle for reannotation thus reducing the overall cost of reannotation. In this setting we tested three existing active learning methods, namely DBAL, Variance of Gradients (VoG) and BADGE. The proposed approach to data denoising is tested on the problem of offensive language detection. We observe that active learning can be effectively used for the purposes of data denoising, however care should be taken when choosing the algorithm for this purpose.", }

Noisily labeled textual data is ample on internet platforms that allow user-created content. Training models, such as offensive language detection models for comment moderation, on such data may prove difficult as the noise in the labels prevents the model to converge. In this work, we propose to use active learning methods for the purposes of denoising training data for model training. The goal is to sample examples the most informative examples with noisy labels with active learning and send them to the oracle for reannotation thus reducing the overall cost of reannotation. In this setting we tested three existing active learning methods, namely DBAL, Variance of Gradients (VoG) and BADGE. The proposed approach to data denoising is tested on the problem of offensive language detection. We observe that active learning can be effectively used for the purposes of data denoising, however care should be taken when choosing the algorithm for this purpose.

Denoising Labeled Data for Comment Moderation Using Active Learning

lrec-main.413

Poster

https://aclanthology.org/2024.lrec-main.414.bib

https://aclanthology.org/2024.lrec-main.414/

@inproceedings{kang-etal-2024-denoising, title = "Denoising Table-Text Retrieval for Open-Domain Question Answering", author = "Kang, Deokhyung and Jung, Baikjin and Kim, Yunsu and Lee, Gary Geunbae", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.414", pages = "4634--4640", abstract = "In table-text open-domain question answering, a retriever system retrieves relevant evidence from tables and text to answer questions. Previous studies in table-text open-domain question answering have two common challenges: firstly, their retrievers can be affected by false-positive labels in training datasets; secondly, they may struggle to provide appropriate evidence for questions that require reasoning across the table. To address these issues, we propose Denoised Table-Text Retriever (DoTTeR). Our approach involves utilizing a denoised training dataset with fewer false positive labels by discarding instances with lower question-relevance scores measured through a false positive detection model. Subsequently, we integrate table-level ranking information into the retriever to assist in finding evidence for questions that demand reasoning across the table. To encode this ranking information, we fine-tune a rank-aware column encoder to identify minimum and maximum values within a column. Experimental results demonstrate that DoTTeR significantly outperforms strong baselines on both retrieval recall and downstream QA tasks. Our code is available at https://github.com/deokhk/DoTTeR.", }

In table-text open-domain question answering, a retriever system retrieves relevant evidence from tables and text to answer questions. Previous studies in table-text open-domain question answering have two common challenges: firstly, their retrievers can be affected by false-positive labels in training datasets; secondly, they may struggle to provide appropriate evidence for questions that require reasoning across the table. To address these issues, we propose Denoised Table-Text Retriever (DoTTeR). Our approach involves utilizing a denoised training dataset with fewer false positive labels by discarding instances with lower question-relevance scores measured through a false positive detection model. Subsequently, we integrate table-level ranking information into the retriever to assist in finding evidence for questions that demand reasoning across the table. To encode this ranking information, we fine-tune a rank-aware column encoder to identify minimum and maximum values within a column. Experimental results demonstrate that DoTTeR significantly outperforms strong baselines on both retrieval recall and downstream QA tasks. Our code is available at https://github.com/deokhk/DoTTeR.

Denoising Table-Text Retrieval for Open-Domain Question Answering

lrec-main.414

Poster

2403.17611

https://aclanthology.org/2024.lrec-main.415.bib

https://aclanthology.org/2024.lrec-main.415/

@inproceedings{luecking-etal-2024-dependencies, title = "Dependencies over Times and Tools ({D}o{TT})", author = "Luecking, Andy and Abrami, Giuseppe and Hammerla, Leon and Rahn, Marc and Baumartz, Daniel and Eger, Steffen and Mehler, Alexander", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.415", pages = "4641--4653", abstract = "Purpose: Based on the examples of English and German, we investigate to what extent parsers trained on modern variants of these languages can be transferred to older language levels without loss. Methods: We developed a treebank called DoTT (https://github.com/texttechnologylab/DoTT) which covers, roughly, the time period from 1800 until today, in conjunction with the further development of the annotation tool DependencyAnnotator. DoTT consists of a collection of diachronic corpora enriched with dependency annotations using 3 parsers, 6 pre-trained language models, 5 newly trained models for German, and two tag sets (TIGER and Universal Dependencies). To assess how the different parsers perform on texts from different time periods, we created a gold standard sample as a benchmark. Results: We found that the parsers/models perform quite well on modern texts (document-level LAS ranging from 82.89 to 88.54) and slightly worse on older texts, as expected (average document-level LAS 84.60 vs. 86.14), but not significantly. For German texts, the (German) TIGER scheme achieved slightly better results than UD. Conclusion: Overall, this result speaks for the transferability of parsers to past language levels, at least dating back until around 1800. This very transferability, it is however argued, means that studies of language change in the field of dependency syntax can draw on dependency distance but miss out on some grammatical phenomena.", }

Purpose: Based on the examples of English and German, we investigate to what extent parsers trained on modern variants of these languages can be transferred to older language levels without loss. Methods: We developed a treebank called DoTT (https://github.com/texttechnologylab/DoTT) which covers, roughly, the time period from 1800 until today, in conjunction with the further development of the annotation tool DependencyAnnotator. DoTT consists of a collection of diachronic corpora enriched with dependency annotations using 3 parsers, 6 pre-trained language models, 5 newly trained models for German, and two tag sets (TIGER and Universal Dependencies). To assess how the different parsers perform on texts from different time periods, we created a gold standard sample as a benchmark. Results: We found that the parsers/models perform quite well on modern texts (document-level LAS ranging from 82.89 to 88.54) and slightly worse on older texts, as expected (average document-level LAS 84.60 vs. 86.14), but not significantly. For German texts, the (German) TIGER scheme achieved slightly better results than UD. Conclusion: Overall, this result speaks for the transferability of parsers to past language levels, at least dating back until around 1800. This very transferability, it is however argued, means that studies of language change in the field of dependency syntax can draw on dependency distance but miss out on some grammatical phenomena.

Dependencies over Times and Tools (DoTT)

lrec-main.415

Poster

https://aclanthology.org/2024.lrec-main.416.bib

https://aclanthology.org/2024.lrec-main.416/

@inproceedings{cao-etal-2024-depth, title = "Depth Aware Hierarchical Replay Continual Learning for Knowledge Based Question Answering", author = "Cao, Zhixiong and Zheng, Hai-Tao and Li, Yangning and Xu, Jin and Li, Rongsheng and Kim, Hong-Gee", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.416", pages = "4654--4664", abstract = "Continual learning is an emerging area of machine learning that deals with the issue where models adapt well to the latest data but lose the ability to remember past data due to changes in the data source. A widely adopted solution is by keeping a small memory of previous learned data that use replay. Most of the previous studies on continual learning focused on classification tasks, such as image classification and text classification, where the model needs only to categorize the input data. Inspired by the human ability to incrementally learn knowledge and solve different problems using learned knowledge, we considered a more pratical scenario, knowledge based quesiton answering about continual learning. In this scenario, each single question is different from others(means different fact trippes to answer them) while classification tasks only need to find feature boundaries of different categories, which are the curves or surfaces that separate different categories in the feature space. To address this issue, we proposed a depth aware hierarchical replay framework which include a tree structure classfier to have a sense of knowledge distribution and fill the gap between text classfication tasks and question-answering tasks for continual learning, a local sampler to grasp these critical samples and a depth aware learning network to reconstructe the feature space of a single learning round. In our experiments, we have demonstrated that our proposed model outperforms previous continual learning methods in mitigating the issue of catastrophic forgetting.", }

Continual learning is an emerging area of machine learning that deals with the issue where models adapt well to the latest data but lose the ability to remember past data due to changes in the data source. A widely adopted solution is by keeping a small memory of previous learned data that use replay. Most of the previous studies on continual learning focused on classification tasks, such as image classification and text classification, where the model needs only to categorize the input data. Inspired by the human ability to incrementally learn knowledge and solve different problems using learned knowledge, we considered a more pratical scenario, knowledge based quesiton answering about continual learning. In this scenario, each single question is different from others(means different fact trippes to answer them) while classification tasks only need to find feature boundaries of different categories, which are the curves or surfaces that separate different categories in the feature space. To address this issue, we proposed a depth aware hierarchical replay framework which include a tree structure classfier to have a sense of knowledge distribution and fill the gap between text classfication tasks and question-answering tasks for continual learning, a local sampler to grasp these critical samples and a depth aware learning network to reconstructe the feature space of a single learning round. In our experiments, we have demonstrated that our proposed model outperforms previous continual learning methods in mitigating the issue of catastrophic forgetting.

Depth Aware Hierarchical Replay Continual Learning for Knowledge Based Question Answering

lrec-main.416

Poster

https://aclanthology.org/2024.lrec-main.417.bib

https://aclanthology.org/2024.lrec-main.417/

@inproceedings{elnokrashy-etal-2024-depth, title = "Depth-Wise Attention ({DWA}tt): A Layer Fusion Method for Data-Efficient Classification", author = "ElNokrashy, Muhammad and AlKhamissi, Badr and Diab, Mona", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.417", pages = "4665--4674", abstract = "Language Models pretrained on large textual data have been shown to encode different types of knowledge simultaneously. Traditionally, only the features from the last layer are used when adapting to new tasks or data. We put forward that, when using or finetuning deep pretrained models, intermediate layer features that may be relevant to the downstream task are buried too deep to be used efficiently in terms of needed samples or steps. To test this, we propose a new layer fusion method: Depth-Wise Attention (DWAtt), to help re-surface signals from non-final layers. We compare DWAtt to a basic concatenation-based layer fusion method (Concat), and compare both to a deeper model baseline{---}all kept within a similar parameter budget. Our findings show that DWAtt and Concat are more step- and sample-efficient than the baseline, especially in the few-shot setting. DWAtt outperforms Concat on larger data sizes. On CoNLL-03 NER, layer fusion shows 3.68 − 9.73{\%} F1 gain at different few-shot sizes. The layer fusion models presented significantly outperform the baseline in various training scenarios with different data sizes, architectures, and training constraints.", }

Language Models pretrained on large textual data have been shown to encode different types of knowledge simultaneously. Traditionally, only the features from the last layer are used when adapting to new tasks or data. We put forward that, when using or finetuning deep pretrained models, intermediate layer features that may be relevant to the downstream task are buried too deep to be used efficiently in terms of needed samples or steps. To test this, we propose a new layer fusion method: Depth-Wise Attention (DWAtt), to help re-surface signals from non-final layers. We compare DWAtt to a basic concatenation-based layer fusion method (Concat), and compare both to a deeper model baseline{---}all kept within a similar parameter budget. Our findings show that DWAtt and Concat are more step- and sample-efficient than the baseline, especially in the few-shot setting. DWAtt outperforms Concat on larger data sizes. On CoNLL-03 NER, layer fusion shows 3.68 − 9.73{\%} F1 gain at different few-shot sizes. The layer fusion models presented significantly outperform the baseline in various training scenarios with different data sizes, architectures, and training constraints.

Depth-Wise Attention (DWAtt): A Layer Fusion Method for Data-Efficient Classification

lrec-main.417

Poster

2209.15168

https://aclanthology.org/2024.lrec-main.418.bib

https://aclanthology.org/2024.lrec-main.418/

@inproceedings{heaton-mitra-2024-deriving, title = "Deriving Entity-Specific Embeddings from Multi-Entity Sequences", author = "Heaton, Connor and Mitra, Prasenjit", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.418", pages = "4675--4684", abstract = "Underpinning much of the recent progress in deep learning is the transformer architecture, which takes as input a sequence of embeddings E and emits an updated sequence of embeddings E{'}. A special [CLS] embedding is often included in this sequence, serving as a description of the sequence once processed and used as the basis for subsequent sequence-level tasks. The processed [CLS] embedding loses utility, however, when the model is presented with a multi-entity sequence and asked to perform an entity-specific task. When processing a multi-speaker dialogue, for example, the [CLS] embedding describes the entire dialogue, not any individual utterance/speaker. Existing methods toward entity-specific prediction involve redundant computation or post-processing outside of the transformer. We present a novel methodology for deriving entity-specific embeddings from a multi-entity sequence completely within the transformer, with a loose definition of entity amenable to many problem spaces. To show the generic applicability of our method, we apply it to widely different tasks: emotion recognition in conversation and player performance projection in baseball and show that it can be used to achieve SOTA in both. Code can be found at https://github.com/c-heat16/EntitySpecificEmbeddings.", }

Underpinning much of the recent progress in deep learning is the transformer architecture, which takes as input a sequence of embeddings E and emits an updated sequence of embeddings E{'}. A special [CLS] embedding is often included in this sequence, serving as a description of the sequence once processed and used as the basis for subsequent sequence-level tasks. The processed [CLS] embedding loses utility, however, when the model is presented with a multi-entity sequence and asked to perform an entity-specific task. When processing a multi-speaker dialogue, for example, the [CLS] embedding describes the entire dialogue, not any individual utterance/speaker. Existing methods toward entity-specific prediction involve redundant computation or post-processing outside of the transformer. We present a novel methodology for deriving entity-specific embeddings from a multi-entity sequence completely within the transformer, with a loose definition of entity amenable to many problem spaces. To show the generic applicability of our method, we apply it to widely different tasks: emotion recognition in conversation and player performance projection in baseball and show that it can be used to achieve SOTA in both. Code can be found at https://github.com/c-heat16/EntitySpecificEmbeddings.

Deriving Entity-Specific Embeddings from Multi-Entity Sequences

lrec-main.418

Poster

https://aclanthology.org/2024.lrec-main.419.bib

https://aclanthology.org/2024.lrec-main.419/

@inproceedings{guo-etal-2024-det, title = "{DET}: A Dual-Encoding Transformer for Relational Graph Embedding", author = "Guo, Lingbing and Chen, Zhuo and Chen, Jiaoyan and Zhang, Qiang and Chen, Huajun", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.419", pages = "4685--4696", abstract = "Despite recent successes in natural language processing and computer vision, Transformer faces scalability issues when processing graphs, e.g., computing the full node-to-node attention on knowledge graphs (KGs) with million of entities is still infeasible. The existing methods mitigate this problem by considering only the local neighbors, sacrificing the Transformer{'}s ability to attend to elements at any distance. This paper proposes a new Transformer architecture called Dual-Encoding Transformer (DET). DET comprises a structural encoder to aggregate information from nearby neighbors, and a semantic encoder to seek for semantically relevant nodes. We adopt a semantic neighbor search approach inspired by multiple sequence alignment (MSA) algorithms used in biological sciences. By stacking the two encoders alternately, similar to the MSA Transformer for protein representation, our method achieves superior performance compared to state-of-the-art attention-based methods on complex relational graphs like KGs and citation networks. Additionally, DET remains competitive for smaller graphs such as molecules.", }

Despite recent successes in natural language processing and computer vision, Transformer faces scalability issues when processing graphs, e.g., computing the full node-to-node attention on knowledge graphs (KGs) with million of entities is still infeasible. The existing methods mitigate this problem by considering only the local neighbors, sacrificing the Transformer{'}s ability to attend to elements at any distance. This paper proposes a new Transformer architecture called Dual-Encoding Transformer (DET). DET comprises a structural encoder to aggregate information from nearby neighbors, and a semantic encoder to seek for semantically relevant nodes. We adopt a semantic neighbor search approach inspired by multiple sequence alignment (MSA) algorithms used in biological sciences. By stacking the two encoders alternately, similar to the MSA Transformer for protein representation, our method achieves superior performance compared to state-of-the-art attention-based methods on complex relational graphs like KGs and citation networks. Additionally, DET remains competitive for smaller graphs such as molecules.

DET: A Dual-Encoding Transformer for Relational Graph Embedding

lrec-main.419

Poster

https://aclanthology.org/2024.lrec-main.420.bib

https://aclanthology.org/2024.lrec-main.420/

@inproceedings{regneri-etal-2024-detecting, title = "Detecting Conceptual Abstraction in {LLM}s", author = "Regneri, Michaela and Abdelhalim, Alhassan and Laue, Soeren", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.420", pages = "4697--4704", abstract = "We show a novel approach to detecting noun abstraction within a large language model (LLM). Starting from a psychologically motivated set of noun pairs in taxonomic relationships, we instantiate surface patterns indicating hypernymy and analyze the attention matrices produced by BERT. We compare the results to two sets of counterfactuals and show that we can detect hypernymy in the abstraction mechanism, which cannot solely be related to the distributional similarity of noun pairs. Our findings are a first step towards the explainability of conceptual abstraction in LLMs.", }

We show a novel approach to detecting noun abstraction within a large language model (LLM). Starting from a psychologically motivated set of noun pairs in taxonomic relationships, we instantiate surface patterns indicating hypernymy and analyze the attention matrices produced by BERT. We compare the results to two sets of counterfactuals and show that we can detect hypernymy in the abstraction mechanism, which cannot solely be related to the distributional similarity of noun pairs. Our findings are a first step towards the explainability of conceptual abstraction in LLMs.

Detecting Conceptual Abstraction in LLMs

lrec-main.420

Poster

2404.15848

https://aclanthology.org/2024.lrec-main.421.bib

https://aclanthology.org/2024.lrec-main.421/

@inproceedings{eo-etal-2024-detecting, title = "Detecting Critical Errors Considering Cross-Cultural Factors in {E}nglish-{K}orean Translation", author = "Eo, Sugyeong and Lim, Jungwoo and Park, Chanjun and Jung, DaHyun and Koo, Seonmin and Moon, Hyeonseok and Seo, Jaehyung and Lim, Heuiseok", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.421", pages = "4705--4716", abstract = "Recent machine translation (MT) systems have overcome language barriers for a wide range of users, yet they still carry the risk of critical meaning deviation. Critical error detection (CED) is a task that identifies an inherent risk of catastrophic meaning distortions in the machine translation output. With the importance of reflecting cultural elements in detecting critical errors, we introduce the culture-aware {``}Politeness{''} type in detecting English-Korean critical translation errors. Besides, we facilitate two tasks by providing multiclass labels: critical error detection and critical error type classification (CETC). Empirical evaluations reveal that our introduced data augmentation approach using a newly presented perturber significantly outperforms existing baselines in both tasks. Further analysis highlights the significance of multiclass labeling by demonstrating its superior effectiveness compared to binary labels.", }

Recent machine translation (MT) systems have overcome language barriers for a wide range of users, yet they still carry the risk of critical meaning deviation. Critical error detection (CED) is a task that identifies an inherent risk of catastrophic meaning distortions in the machine translation output. With the importance of reflecting cultural elements in detecting critical errors, we introduce the culture-aware {``}Politeness{''} type in detecting English-Korean critical translation errors. Besides, we facilitate two tasks by providing multiclass labels: critical error detection and critical error type classification (CETC). Empirical evaluations reveal that our introduced data augmentation approach using a newly presented perturber significantly outperforms existing baselines in both tasks. Further analysis highlights the significance of multiclass labeling by demonstrating its superior effectiveness compared to binary labels.

Detecting Critical Errors Considering Cross-Cultural Factors in English-Korean Translation

lrec-main.421

Poster

https://aclanthology.org/2024.lrec-main.422.bib

https://aclanthology.org/2024.lrec-main.422/

@inproceedings{ullah-etal-2024-detecting, title = "Detecting Cybercrimes in Accordance with {P}akistani Law: Dataset and Evaluation Using {PLM}s", author = "Ullah, Faizad and Faheem, Ali and Azam, Ubaid and Ayub, Muhammad Sohaib and Kamiran, Faisal and Karim, Asim", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.422", pages = "4717--4728", abstract = "Cybercrime is a serious and growing threat affecting millions of people worldwide. Detecting cybercrimes from text messages is challenging, as it requires understanding the linguistic and cultural nuances of different languages and regions. Roman Urdu is a widely used language in Pakistan and other South Asian countries, however, it lacks sufficient resources and tools for natural language processing and cybercrime detection. To address this problem, we make three main contributions in this paper. (1) We create and release CRU, a benchmark dataset for text-based cybercrime detection in Roman Urdu, which covers a number of cybercrimes as defined by the Prevention of Electronic Crimes Act (PECA) of Pakistan. This dataset is annotated by experts following a standardized procedure based on Pakistan{'}s legal framework. (2) We perform experiments on four pre-trained language models (PLMs) for cybercrime text classification in Roman Urdu. Our results show that xlm-roberta-base is the best model for this task, achieving the highest performance on all metrics. (3) We explore the utility of prompt engineering techniques, namely prefix and cloze prompts, for enhancing the performance of PLMs for low-resource languages such as Roman Urdu. We analyze the impact of different prompt shapes and $k$-shot settings on the performance of xlm-roberta-base and bert-base-multilingual-cased. We find that prefix prompts are more effective than cloze prompts for Roman Urdu classification tasks, as they provide more contextually relevant completions for the models. Our work provides useful insights and resources for future research on cybercrime detection and text classification in low-resource languages.", }

Cybercrime is a serious and growing threat affecting millions of people worldwide. Detecting cybercrimes from text messages is challenging, as it requires understanding the linguistic and cultural nuances of different languages and regions. Roman Urdu is a widely used language in Pakistan and other South Asian countries, however, it lacks sufficient resources and tools for natural language processing and cybercrime detection. To address this problem, we make three main contributions in this paper. (1) We create and release CRU, a benchmark dataset for text-based cybercrime detection in Roman Urdu, which covers a number of cybercrimes as defined by the Prevention of Electronic Crimes Act (PECA) of Pakistan. This dataset is annotated by experts following a standardized procedure based on Pakistan{'}s legal framework. (2) We perform experiments on four pre-trained language models (PLMs) for cybercrime text classification in Roman Urdu. Our results show that xlm-roberta-base is the best model for this task, achieving the highest performance on all metrics. (3) We explore the utility of prompt engineering techniques, namely prefix and cloze prompts, for enhancing the performance of PLMs for low-resource languages such as Roman Urdu. We analyze the impact of different prompt shapes and $k$-shot settings on the performance of xlm-roberta-base and bert-base-multilingual-cased. We find that prefix prompts are more effective than cloze prompts for Roman Urdu classification tasks, as they provide more contextually relevant completions for the models. Our work provides useful insights and resources for future research on cybercrime detection and text classification in low-resource languages.

Detecting Cybercrimes in Accordance with Pakistani Law: Dataset and Evaluation Using PLMs

lrec-main.422

Poster

https://aclanthology.org/2024.lrec-main.423.bib

https://aclanthology.org/2024.lrec-main.423/

@inproceedings{chang-etal-2024-detecting, title = "Detecting Hallucination and Coverage Errors in Retrieval Augmented Generation for Controversial Topics", author = "Chang, Tyler A. and Tomanek, Katrin and Hoffmann, Jessica and Thain, Nithum and MacMurray van Liemt, Erin and Meier-Hellstern, Kathleen and Dixon, Lucas", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.423", pages = "4729--4743", abstract = "We explore a strategy to handle controversial topics in LLM-based chatbots based on Wikipedia{'}s Neutral Point of View (NPOV) principle: acknowledge the absence of a single true answer and surface multiple perspectives. We frame this as retrieval augmented generation, where perspectives are retrieved from a knowledge base and the LLM is tasked with generating a fluent and faithful response from the given perspectives. As a starting point, we use a deterministic retrieval system and then focus on common LLM failure modes that arise during this approach to text generation, namely hallucination and coverage errors. We propose and evaluate three methods to detect such errors based on (1) word-overlap, (2) salience, and (3) LLM-based classifiers. Our results demonstrate that LLM-based classifiers, even when trained only on synthetic errors, achieve high error detection performance, with ROC AUC scores of 95.3{\%} for hallucination and 90.5{\%} for coverage error detection on unambiguous error cases. We show that when no training data is available, our other methods still yield good results on hallucination (84.0{\%}) and coverage error (85.2{\%}) detection.", }

We explore a strategy to handle controversial topics in LLM-based chatbots based on Wikipedia{'}s Neutral Point of View (NPOV) principle: acknowledge the absence of a single true answer and surface multiple perspectives. We frame this as retrieval augmented generation, where perspectives are retrieved from a knowledge base and the LLM is tasked with generating a fluent and faithful response from the given perspectives. As a starting point, we use a deterministic retrieval system and then focus on common LLM failure modes that arise during this approach to text generation, namely hallucination and coverage errors. We propose and evaluate three methods to detect such errors based on (1) word-overlap, (2) salience, and (3) LLM-based classifiers. Our results demonstrate that LLM-based classifiers, even when trained only on synthetic errors, achieve high error detection performance, with ROC AUC scores of 95.3{\%} for hallucination and 90.5{\%} for coverage error detection on unambiguous error cases. We show that when no training data is available, our other methods still yield good results on hallucination (84.0{\%}) and coverage error (85.2{\%}) detection.

Detecting Hallucination and Coverage Errors in Retrieval Augmented Generation for Controversial Topics

lrec-main.423

Poster

2403.08904

https://aclanthology.org/2024.lrec-main.424.bib

https://aclanthology.org/2024.lrec-main.424/

@inproceedings{becker-etal-2024-detecting, title = "Detecting Impact Relevant Sections in Scientific Research", author = "Becker, Maria and Han, Kanyao and Lee, Haejin and Werthmann, Antonina and Rezapour, Rezvaneh and Diesner, Jana and Witt, Andreas", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.424", pages = "4744--4749", abstract = "Impact assessment is an evolving area of research that aims at measuring and predicting the potential effects of projects or programs. Measuring the impact of scientific research is a vibrant subdomain, closely intertwined with impact assessment. A recurring obstacle pertains to the absence of an efficient framework which can facilitate the analysis of lengthy reports and text labeling. To address this issue, we propose a framework for automatically assessing the impact of scientific research projects by identifying pertinent sections in project reports that indicate the potential impacts. We leverage a mixed-method approach, combining manual annotations with supervised machine learning, to extract these passages from project reports. We experiment with different machine learning algorithms, including traditional statistical models as well as pre-trained transformer language models. Our experiments show that our proposed method achieves accuracy scores up to 0.81, and that our method is generalizable to scientific research from different domains and different languages.", }

Impact assessment is an evolving area of research that aims at measuring and predicting the potential effects of projects or programs. Measuring the impact of scientific research is a vibrant subdomain, closely intertwined with impact assessment. A recurring obstacle pertains to the absence of an efficient framework which can facilitate the analysis of lengthy reports and text labeling. To address this issue, we propose a framework for automatically assessing the impact of scientific research projects by identifying pertinent sections in project reports that indicate the potential impacts. We leverage a mixed-method approach, combining manual annotations with supervised machine learning, to extract these passages from project reports. We experiment with different machine learning algorithms, including traditional statistical models as well as pre-trained transformer language models. Our experiments show that our proposed method achieves accuracy scores up to 0.81, and that our method is generalizable to scientific research from different domains and different languages.

Detecting Impact Relevant Sections in Scientific Research

lrec-main.424

Poster

https://aclanthology.org/2024.lrec-main.425.bib

https://aclanthology.org/2024.lrec-main.425/

@inproceedings{ali-etal-2024-detecting, title = "Detecting Loanwords in Emakhuwa: An Extremely Low-Resource {B}antu Language Exhibiting Significant Borrowing from {P}ortuguese", author = "Ali, Felermino Dario Mario and Lopes Cardoso, Henrique and Sousa-Silva, Rui", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.425", pages = "4750--4759", abstract = "The accurate identification of loanwords within a given text holds significant potential as a valuable tool for addressing data augmentation and mitigating data sparsity issues. Such identification can improve the performance of various natural language processing tasks, particularly in the context of low-resource languages that lack standardized spelling conventions.This research proposes a supervised method to identify loanwords in Emakhuwa, borrowed from Portuguese. Our methodology encompasses a two-fold approach. Firstly, we employ traditional machine learning algorithms incorporating handcrafted features, including language-specific and similarity-based features. We build upon prior studies to extract similarity features and propose utilizing two external resources: a Sequence-to-Sequence model and a dictionary. This innovative approach allows us to identify loanwords solely by analyzing the target word without prior knowledge about its donor counterpart. Furthermore, we fine-tune the pre-trained CANINE model for the downstream task of loanword detection, which culminates in the impressive achievement of the F1-score of 93{\%}. To the best of our knowledge, this study is the first of its kind focusing on Emakhuwa, and the preliminary results are promising as they pave the way to further advancements.", }

The accurate identification of loanwords within a given text holds significant potential as a valuable tool for addressing data augmentation and mitigating data sparsity issues. Such identification can improve the performance of various natural language processing tasks, particularly in the context of low-resource languages that lack standardized spelling conventions.This research proposes a supervised method to identify loanwords in Emakhuwa, borrowed from Portuguese. Our methodology encompasses a two-fold approach. Firstly, we employ traditional machine learning algorithms incorporating handcrafted features, including language-specific and similarity-based features. We build upon prior studies to extract similarity features and propose utilizing two external resources: a Sequence-to-Sequence model and a dictionary. This innovative approach allows us to identify loanwords solely by analyzing the target word without prior knowledge about its donor counterpart. Furthermore, we fine-tune the pre-trained CANINE model for the downstream task of loanword detection, which culminates in the impressive achievement of the F1-score of 93{\%}. To the best of our knowledge, this study is the first of its kind focusing on Emakhuwa, and the preliminary results are promising as they pave the way to further advancements.

Detecting Loanwords in Emakhuwa: An Extremely Low-Resource Bantu Language Exhibiting Significant Borrowing from Portuguese

lrec-main.425

Poster

https://aclanthology.org/2024.lrec-main.426.bib

https://aclanthology.org/2024.lrec-main.426/

@inproceedings{song-etal-2024-detecting, title = "Detecting Offensive Language in an Open Chatbot Platform", author = "Song, Hyeonho and Hong, Jisu and Jung, Chani and Chin, Hyojin and Shin, Mingi and Choi, Yubin and Choi, Junghoi and Cha, Meeyoung", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.426", pages = "4760--4771", abstract = "While detecting offensive language in online spaces remains an important societal issue, there is still a significant gap in existing research and practial datasets specific to chatbots. Furthermore, many of the current efforts by service providers to automatically filter offensive language are vulnerable to users{'} deliberate text manipulation tactics, such as misspelling words. In this study, we analyze offensive language patterns in real logs of 6,254,261 chat utterance pairs from the commercial chat service Simsimi, which cover a variety of conversation topics. Based on the observed patterns, we introduce a novel offensive language detection method{---}a contrastive learning model that embeds chat content with a random masking strategy. We show that this model outperforms existing models in detecting offensive language in open-domain chat conversations while also demonstrating robustness against users{'} deliberate text manipulation tactics when using offensive language. We release our curated chatbot dataset to foster research on offensive language detection in open-domain conversations and share lessons learned from mitigating offensive language on a live platform.", }

While detecting offensive language in online spaces remains an important societal issue, there is still a significant gap in existing research and practial datasets specific to chatbots. Furthermore, many of the current efforts by service providers to automatically filter offensive language are vulnerable to users{'} deliberate text manipulation tactics, such as misspelling words. In this study, we analyze offensive language patterns in real logs of 6,254,261 chat utterance pairs from the commercial chat service Simsimi, which cover a variety of conversation topics. Based on the observed patterns, we introduce a novel offensive language detection method{---}a contrastive learning model that embeds chat content with a random masking strategy. We show that this model outperforms existing models in detecting offensive language in open-domain chat conversations while also demonstrating robustness against users{'} deliberate text manipulation tactics when using offensive language. We release our curated chatbot dataset to foster research on offensive language detection in open-domain conversations and share lessons learned from mitigating offensive language on a live platform.

Detecting Offensive Language in an Open Chatbot Platform

lrec-main.426

Poster

https://aclanthology.org/2024.lrec-main.427.bib

https://aclanthology.org/2024.lrec-main.427/

@inproceedings{clerice-2024-detecting, title = "Detecting Sexual Content at the Sentence Level in First Millennium {L}atin Texts", author = "Clerice, Thibault", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.427", pages = "4772--4783", abstract = "In this study, we propose to evaluate the use of deep learning methods for semantic classification at the sentence level to accelerate the process of corpus building in the field of humanities and linguistics, a traditional and time-consuming task. We introduce a novel corpus comprising around 2500 sentences spanning from 300 BCE to 900 CE including sexual semantics (medical, erotica, etc.). We evaluate various sentence classification approaches and different input embedding layers, and show that all consistently outperform simple token-based searches. We explore the integration of idiolectal and sociolectal metadata embeddings (centuries, author, type of writing), but find that it leads to overfitting. Our results demonstrate the effectiveness of this approach, achieving high precision and true positive rates (TPR) of respectively 70.60{\%} and 86.33{\%} using HAN. We evaluate the impact of the dataset size on the model performances (420 instead of 2013 training samples), and show that, while our models perform worse, they still offer a high enough precision and TPR, even without MLM, respectively 69{\%} and 51{\%}. Given the result, we provide an analysis of the attention mechanism as a supporting added value for humanists in order to produce more data.", }

In this study, we propose to evaluate the use of deep learning methods for semantic classification at the sentence level to accelerate the process of corpus building in the field of humanities and linguistics, a traditional and time-consuming task. We introduce a novel corpus comprising around 2500 sentences spanning from 300 BCE to 900 CE including sexual semantics (medical, erotica, etc.). We evaluate various sentence classification approaches and different input embedding layers, and show that all consistently outperform simple token-based searches. We explore the integration of idiolectal and sociolectal metadata embeddings (centuries, author, type of writing), but find that it leads to overfitting. Our results demonstrate the effectiveness of this approach, achieving high precision and true positive rates (TPR) of respectively 70.60{\%} and 86.33{\%} using HAN. We evaluate the impact of the dataset size on the model performances (420 instead of 2013 training samples), and show that, while our models perform worse, they still offer a high enough precision and TPR, even without MLM, respectively 69{\%} and 51{\%}. Given the result, we provide an analysis of the attention mechanism as a supporting added value for humanists in order to produce more data.

Detecting Sexual Content at the Sentence Level in First Millennium Latin Texts

lrec-main.427

Poster

2309.14974

https://aclanthology.org/2024.lrec-main.428.bib

https://aclanthology.org/2024.lrec-main.428/

@inproceedings{dou-etal-2024-detection, title = "Detection, Diagnosis, and Explanation: A Benchmark for {C}hinese Medial Hallucination Evaluation", author = "Dou, Chengfeng and Zhang, Ying and Chen, Yanyuan and Jin, Zhi and Jiao, Wenpin and Zhao, Haiyan and Huang, Yu", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.428", pages = "4784--4794", abstract = "Large Language Models (LLMs) have made significant progress recently. However, their practical use in healthcare is hindered by their tendency to generate hallucinations. One specific type, called snowballing hallucination, occurs when LLMs encounter misleading information, and poses a security threat to LLMs. To understand how well LLMs can resist these hallucination, we create the Chinese Medical Hallucination Evaluation benchmark (CMHE). This benchmark can be used to evaluate LLMs{'} ability to detect medical hallucinations, make accurate diagnoses in noisy conditions, and provide plausible explanations. The creation of this benchmark involves a combination of manual and model-based approaches. In addition, we use ICD-10 as well as MeSH, two specialized glossaries, to aid in the evaluation. Our experiments show that the LLM struggles to identify fake medical terms and makes poor diagnoses in distracting environments. However, improving the model{'}s understanding of medical concepts can help it resist interference to some extent.", }

Large Language Models (LLMs) have made significant progress recently. However, their practical use in healthcare is hindered by their tendency to generate hallucinations. One specific type, called snowballing hallucination, occurs when LLMs encounter misleading information, and poses a security threat to LLMs. To understand how well LLMs can resist these hallucination, we create the Chinese Medical Hallucination Evaluation benchmark (CMHE). This benchmark can be used to evaluate LLMs{'} ability to detect medical hallucinations, make accurate diagnoses in noisy conditions, and provide plausible explanations. The creation of this benchmark involves a combination of manual and model-based approaches. In addition, we use ICD-10 as well as MeSH, two specialized glossaries, to aid in the evaluation. Our experiments show that the LLM struggles to identify fake medical terms and makes poor diagnoses in distracting environments. However, improving the model{'}s understanding of medical concepts can help it resist interference to some extent.

Detection, Diagnosis, and Explanation: A Benchmark for Chinese Medial Hallucination Evaluation

lrec-main.428

Poster

https://aclanthology.org/2024.lrec-main.429.bib

https://aclanthology.org/2024.lrec-main.429/

@inproceedings{oneil-etal-2024-developing, title = "Developing a Benchmark for Pronunciation Feedback: Creation of a Phonemically Annotated Speech Corpus of isi{Z}ulu Language Learner Speech", author = "O{'}Neil, Alexandra and Hjortnaes, Nils and Tyers, Francis and Nkosi, Zinhle and Ndlovu, Thulile and Mlondo, Zanele and Pewa, Ngami Phumzile", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.429", pages = "4795--4801", abstract = "Pronunciation of the phonemic inventory of a new language often presents difficulties to second language (L2) learners. These challenges can be alleviated by the development of pronunciation feedback tools that take speech input from learners and return information about errors in the utterance. This paper presents the development of a corpus designed for use in pronunciation feedback research. The corpus is comprised of gold standard recordings from isiZulu teachers and recordings from isiZulu L2 learners that have been annotated for pronunciation errors. Exploring the potential benefits of word-level versus phoneme-level feedback necessitates a speech corpus that has been annotated for errors on the phoneme-level. To aid in this discussion, this corpus of isiZulu L2 speech has been annotated for phoneme-errors in utterances, as well as suprasegmental errors in tone.", }

Pronunciation of the phonemic inventory of a new language often presents difficulties to second language (L2) learners. These challenges can be alleviated by the development of pronunciation feedback tools that take speech input from learners and return information about errors in the utterance. This paper presents the development of a corpus designed for use in pronunciation feedback research. The corpus is comprised of gold standard recordings from isiZulu teachers and recordings from isiZulu L2 learners that have been annotated for pronunciation errors. Exploring the potential benefits of word-level versus phoneme-level feedback necessitates a speech corpus that has been annotated for errors on the phoneme-level. To aid in this discussion, this corpus of isiZulu L2 speech has been annotated for phoneme-errors in utterances, as well as suprasegmental errors in tone.

Developing a Benchmark for Pronunciation Feedback: Creation of a Phonemically Annotated Speech Corpus of isiZulu Language Learner Speech

lrec-main.429

Poster

https://aclanthology.org/2024.lrec-main.430.bib

https://aclanthology.org/2024.lrec-main.430/

@inproceedings{polakova-etal-2024-developing, title = "Developing a {R}hetorical {S}tructure {T}heory Treebank for {C}zech", author = "Polakova, Lucie and M{\'\i}rovsk{\'y}, Ji{\v{r}}{\'\i} and Zik{\'a}nov{\'a}, {\v{S}}{\'a}rka and Hajicova, Eva", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.430", pages = "4802--4810", abstract = "We introduce the first version of the Czech RST Discourse Treebank, a collection of Czech journalistic texts manually annotated using the Rhetorical Structure Theory (RST), a global coherence model proposed by Mann and Thompson (1988). Each document in the corpus is represented as a single tree-like structure, where discourse units are interconnected through hierarchical rhetorical relations and their relative importance for the main purpose of a text is modeled by the nuclearity principle. The treebank is freely available in the LINDAT/CLARIAH-CZ repository under the Creative Commons license; for some documents, it includes two gold annotations representing divergent yet relevant interpretations. The paper outlines the annotation process, provides corpus statistics and evaluation, and discusses the issue of consistency associated with the global level of textual interpretation. In general, good agreement on the structure and labeling could be achieved on the lowest, local tree level and on the identification of the most central (nuclear) elementary discourse units. Disagreements mostly concerned segmentation and, in the structure, differences in the stepwise process of linking the largest text blocks. The project contributes to the advancement of RST research and its application to real-world text analysis challenges.", }

We introduce the first version of the Czech RST Discourse Treebank, a collection of Czech journalistic texts manually annotated using the Rhetorical Structure Theory (RST), a global coherence model proposed by Mann and Thompson (1988). Each document in the corpus is represented as a single tree-like structure, where discourse units are interconnected through hierarchical rhetorical relations and their relative importance for the main purpose of a text is modeled by the nuclearity principle. The treebank is freely available in the LINDAT/CLARIAH-CZ repository under the Creative Commons license; for some documents, it includes two gold annotations representing divergent yet relevant interpretations. The paper outlines the annotation process, provides corpus statistics and evaluation, and discusses the issue of consistency associated with the global level of textual interpretation. In general, good agreement on the structure and labeling could be achieved on the lowest, local tree level and on the identification of the most central (nuclear) elementary discourse units. Disagreements mostly concerned segmentation and, in the structure, differences in the stepwise process of linking the largest text blocks. The project contributes to the advancement of RST research and its application to real-world text analysis challenges.

Developing a Rhetorical Structure Theory Treebank for Czech

lrec-main.430

Poster

https://aclanthology.org/2024.lrec-main.431.bib

https://aclanthology.org/2024.lrec-main.431/

@inproceedings{al-laith-etal-2024-development, title = "Development and Evaluation of Pre-trained Language Models for Historical {D}anish and {N}orwegian Literary Texts", author = "Al-Laith, Ali and Conroy, Alexander and Bjerring-Hansen, Jens and Hershcovich, Daniel", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.431", pages = "4811--4819", abstract = "We develop and evaluate the first pre-trained language models specifically tailored for historical Danish and Norwegian texts. Three models are trained on a corpus of 19th-century Danish and Norwegian literature: two directly on the corpus with no prior pre-training, and one with continued pre-training. To evaluate the models, we utilize an existing sentiment classification dataset, and additionally introduce a new annotated word sense disambiguation dataset focusing on the concept of fate. Our assessment reveals that the model employing continued pre-training outperforms the others in two downstream NLP tasks on historical texts. Specifically, we observe substantial improvement in sentiment classification and word sense disambiguation compared to models trained on contemporary texts. These results highlight the effectiveness of continued pre-training for enhancing performance across various NLP tasks in historical text analysis.", }

We develop and evaluate the first pre-trained language models specifically tailored for historical Danish and Norwegian texts. Three models are trained on a corpus of 19th-century Danish and Norwegian literature: two directly on the corpus with no prior pre-training, and one with continued pre-training. To evaluate the models, we utilize an existing sentiment classification dataset, and additionally introduce a new annotated word sense disambiguation dataset focusing on the concept of fate. Our assessment reveals that the model employing continued pre-training outperforms the others in two downstream NLP tasks on historical texts. Specifically, we observe substantial improvement in sentiment classification and word sense disambiguation compared to models trained on contemporary texts. These results highlight the effectiveness of continued pre-training for enhancing performance across various NLP tasks in historical text analysis.

Development and Evaluation of Pre-trained Language Models for Historical Danish and Norwegian Literary Texts

lrec-main.431

Poster

https://aclanthology.org/2024.lrec-main.432.bib

https://aclanthology.org/2024.lrec-main.432/

@inproceedings{james-etal-2024-development, title = "Development of Community-Oriented Text-to-Speech Models for {M}{\=a}ori {`}Avaiki {N}ui ({C}ook {I}slands {M}{\=a}ori)", author = "James, Jesin and Coto-Solano, Rolando and Nicholas, Sally Akevai and Zhu, Joshua and Yu, Bovey and Babasaki, Fuki and Wang, Jenny Tyler and Derby, Nicholas", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.432", pages = "4820--4831", abstract = {In this paper we describe the development of a text-to-speech system for M{\=a}ori {`}Avaiki Nui (Cook Islands M{\=a}ori). We provide details about the process of community-collaboration that was followed throughout the project, a continued engagement where we are trying to develop speech and language technology for the benefit of the community. During this process we gathered a group of recordings that we used to train a TTS system. When training we used two approaches, the HMM-system MaryTTS (Schr{\"o}der et al., 2011) and the deep learning system FastSpeech2 (Ren et al., 2020). We performed two evaluation tasks on the models: First, we measured their quality by having the synthesized speech transcribed by ASR. The human produced ground truth had lower error rates (CER=4.3, WER=18), but the FastSpeech2 audio has lower error rates (CER=11.8 and WER=42.7) than the MaryTTS voice (CER=17.9 and WER=48.1). The second evaluation was a survey amongst speakers of the language so they could judge the voice{'}s quality. The ground truth was rated with the highest quality (MOS=4.6), but the FastSpeech2 voice had an overall quality of MOS=3.2, which was significantly higher than that of the MaryTTS synthesized recordings (MOS=2.0). We intend to use the FastSpeech2 model to create language learning tools for community members both on the Cook Islands and in the diaspora.}, }

In this paper we describe the development of a text-to-speech system for M{\=a}ori {`}Avaiki Nui (Cook Islands M{\=a}ori). We provide details about the process of community-collaboration that was followed throughout the project, a continued engagement where we are trying to develop speech and language technology for the benefit of the community. During this process we gathered a group of recordings that we used to train a TTS system. When training we used two approaches, the HMM-system MaryTTS (Schr{\"o}der et al., 2011) and the deep learning system FastSpeech2 (Ren et al., 2020). We performed two evaluation tasks on the models: First, we measured their quality by having the synthesized speech transcribed by ASR. The human produced ground truth had lower error rates (CER=4.3, WER=18), but the FastSpeech2 audio has lower error rates (CER=11.8 and WER=42.7) than the MaryTTS voice (CER=17.9 and WER=48.1). The second evaluation was a survey amongst speakers of the language so they could judge the voice{'}s quality. The ground truth was rated with the highest quality (MOS=4.6), but the FastSpeech2 voice had an overall quality of MOS=3.2, which was significantly higher than that of the MaryTTS synthesized recordings (MOS=2.0). We intend to use the FastSpeech2 model to create language learning tools for community members both on the Cook Islands and in the diaspora.

Development of Community-Oriented Text-to-Speech Models for Māori `Avaiki Nui (Cook Islands Māori)

lrec-main.432

Poster

https://aclanthology.org/2024.lrec-main.433.bib

https://aclanthology.org/2024.lrec-main.433/

@inproceedings{ning-etal-2024-dgot, title = "{DG}o{T}: Dynamic Graph of Thoughts for Scientific Abstract Generation", author = "Ning, Xinyu and Zhao, Yutong and Liu, Yitong and Yang, Hongwen", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.433", pages = "4832--4846", abstract = "The method of training language models based on domain datasets has obtained significant achievements in the task of generating scientific paper abstracts. However, such models face problems of generalization and expensive training costs. The use of large language models (LLMs) to solve the task of generating paper abstracts saves the cost of model training. However, due to the hallucination problem of LLM, it is often necessary to improve the reliability of the results through multi-round query prompt approach such as Graph of Thoughts (GoT), which also brings additional reasoning costs. In this paper, we propose a Dynamic Graph of Thought (DGoT). It not only inherits the advantages of the existing GoT prompt approach, but also dynamically adjust the graph structure according to data characteristics while reducing model reasoning cost. Experimental results show that our method{'}s cost-effectiveness in abstract generation tasks is only 43.7{\%} to 56.4{\%} of other multi-round query prompt approaches. Our code is available at https://github.com/JayceNing/DGoT.", }

The method of training language models based on domain datasets has obtained significant achievements in the task of generating scientific paper abstracts. However, such models face problems of generalization and expensive training costs. The use of large language models (LLMs) to solve the task of generating paper abstracts saves the cost of model training. However, due to the hallucination problem of LLM, it is often necessary to improve the reliability of the results through multi-round query prompt approach such as Graph of Thoughts (GoT), which also brings additional reasoning costs. In this paper, we propose a Dynamic Graph of Thought (DGoT). It not only inherits the advantages of the existing GoT prompt approach, but also dynamically adjust the graph structure according to data characteristics while reducing model reasoning cost. Experimental results show that our method{'}s cost-effectiveness in abstract generation tasks is only 43.7{\%} to 56.4{\%} of other multi-round query prompt approaches. Our code is available at https://github.com/JayceNing/DGoT.

DGoT: Dynamic Graph of Thoughts for Scientific Abstract Generation

lrec-main.433

Poster

2403.17491

https://aclanthology.org/2024.lrec-main.434.bib

https://aclanthology.org/2024.lrec-main.434/

@inproceedings{nunnari-etal-2024-dgs, title = "{DGS}-Fabeln-1: A Multi-Angle Parallel Corpus of Fairy Tales between {G}erman {S}ign {L}anguage and {G}erman Text", author = "Nunnari, Fabrizio and Avramidis, Eleftherios and Espa{\~n}a-Bonet, Cristina and Gonz{\'a}lez, Marco and Hennes, Anna and Gebhard, Patrick", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.434", pages = "4847--4857", abstract = "We present the acquisition process and the data of DGS-Fabeln-1, a parallel corpus of German text and videos containing German fairy tales interpreted into the German Sign Language (DGS) by a native DGS signer. The corpus contains 573 segments of videos with a total duration of 1 hour and 32 minutes, corresponding with 1428 written sentences. It is the first corpus of semi-naturally expressed DGS that has been filmed from 7 angles, and one of the few sign language (SL) corpora globally which have been filmed from more than 3 angles and where the listener has been simultaneously filmed. The corpus aims at aiding research at SL linguistics, SL machine translation and affective computing, and is freely available for research purposes at the following address: https://doi.org/10.5281/zenodo.10822097.", }

We present the acquisition process and the data of DGS-Fabeln-1, a parallel corpus of German text and videos containing German fairy tales interpreted into the German Sign Language (DGS) by a native DGS signer. The corpus contains 573 segments of videos with a total duration of 1 hour and 32 minutes, corresponding with 1428 written sentences. It is the first corpus of semi-naturally expressed DGS that has been filmed from 7 angles, and one of the few sign language (SL) corpora globally which have been filmed from more than 3 angles and where the listener has been simultaneously filmed. The corpus aims at aiding research at SL linguistics, SL machine translation and affective computing, and is freely available for research purposes at the following address: https://doi.org/10.5281/zenodo.10822097.

DGS-Fabeln-1: A Multi-Angle Parallel Corpus of Fairy Tales between German Sign Language and German Text

lrec-main.434

Poster

https://aclanthology.org/2024.lrec-main.435.bib

https://aclanthology.org/2024.lrec-main.435/

@inproceedings{mizumoto-etal-2024-dialogue, title = "Dialogue Systems Can Generate Appropriate Responses without the Use of Question Marks?{--} a Study of the Effects of {``}?{''} for Spoken Dialogue Systems {--}", author = "Mizumoto, Tomoya and Yamazaki, Takato and Yoshikawa, Katsumasa and Ohagi, Masaya and Kawamoto, Toshiki and Sato, Toshinori", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.435", pages = "4858--4864", abstract = "When individuals engage in spoken discourse, various phenomena can be observed that differ from those that are apparent in text-based conversation. While written communication commonly uses a question mark to denote a query, in spoken discourse, queries are frequently indicated by a rising intonation at the end of a sentence. However, numerous speech recognition engines do not append a question mark to recognized queries, presenting a challenge when creating a spoken dialogue system. Specifically, the absence of a question mark at the end of a sentence can impede the generation of appropriate responses to queries in spoken dialogue systems. Hence, we investigate the impact of question marks on dialogue systems, with the results showing that they have a significant impact. Moreover, we analyze specific examples in an effort to determine which types of utterances have the impact on dialogue systems.", }

When individuals engage in spoken discourse, various phenomena can be observed that differ from those that are apparent in text-based conversation. While written communication commonly uses a question mark to denote a query, in spoken discourse, queries are frequently indicated by a rising intonation at the end of a sentence. However, numerous speech recognition engines do not append a question mark to recognized queries, presenting a challenge when creating a spoken dialogue system. Specifically, the absence of a question mark at the end of a sentence can impede the generation of appropriate responses to queries in spoken dialogue systems. Hence, we investigate the impact of question marks on dialogue systems, with the results showing that they have a significant impact. Moreover, we analyze specific examples in an effort to determine which types of utterances have the impact on dialogue systems.

Dialogue Systems Can Generate Appropriate Responses without the Use of Question Marks?– a Study of the Effects of “?” for Spoken Dialogue Systems –

lrec-main.435

Poster

https://aclanthology.org/2024.lrec-main.436.bib

https://aclanthology.org/2024.lrec-main.436/

@inproceedings{israeli-etal-2024-diaset, title = "{D}ia{S}et: An Annotated Dataset of {A}rabic Conversations", author = "Israeli, Abraham and Naaman, Aviv and Maduel, Guy and Makhoul, Rawaa and Qaraeen, Dana and Ejmail, Amir and Lisnanskey, Dina and Jubran, Julian and Fine, Shai and Bar, Kfir", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.436", pages = "4865--4876", abstract = "We introduce DiaSet, a novel dataset of dialectical Arabic speech, manually transcribed and annotated for two specific downstream tasks: sentiment analysis and named entity recognition. The dataset encapsulates the Palestine dialect, predominantly spoken in Palestine, Israel, and Jordan. Our dataset incorporates authentic conversations between YouTube influencers and their respective guests. Furthermore, we have enriched the dataset with simulated conversations initiated by inviting participants from various locales within the said regions. The participants were encouraged to engage in dialogues with our interviewer. Overall, DiaSet consists of 644.8K tokens and 23.2K annotated instances. Uniform writing standards were upheld during the transcription process. Additionally, we established baseline models by leveraging some of the pre-existing Arabic BERT language models, showcasing the potential applications and efficiencies of our dataset. We make DiaSet publicly available for further research.", }

We introduce DiaSet, a novel dataset of dialectical Arabic speech, manually transcribed and annotated for two specific downstream tasks: sentiment analysis and named entity recognition. The dataset encapsulates the Palestine dialect, predominantly spoken in Palestine, Israel, and Jordan. Our dataset incorporates authentic conversations between YouTube influencers and their respective guests. Furthermore, we have enriched the dataset with simulated conversations initiated by inviting participants from various locales within the said regions. The participants were encouraged to engage in dialogues with our interviewer. Overall, DiaSet consists of 644.8K tokens and 23.2K annotated instances. Uniform writing standards were upheld during the transcription process. Additionally, we established baseline models by leveraging some of the pre-existing Arabic BERT language models, showcasing the potential applications and efficiencies of our dataset. We make DiaSet publicly available for further research.

DiaSet: An Annotated Dataset of Arabic Conversations

lrec-main.436

Poster

https://aclanthology.org/2024.lrec-main.437.bib

https://aclanthology.org/2024.lrec-main.437/

@inproceedings{maes-etal-2024-get, title = "Did You Get It? A Zero-Shot Approach to Locate Information Transfers in Conversations", author = {Ma{\"e}s, Eliot and Boudraa, Hossam and Blache, Philippe and Becerra-Bonache, Leonor}, editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.437", pages = "4877--4890", abstract = "Interaction theories suggest that the emergence of mutual understanding between speakers in natural conversations depends on the construction of a shared knowledge base (\textit{common ground}), but the details of which information and the circumstances under which it is memorized are not explained by any model. Previous works have looked at metrics derived from Information Theory to quantify the dynamics of information exchanged between participants, but do not provide an efficient way to locate information that will enter the common ground. We propose a new method based on the segmentation of a conversation into themes followed by their summarization. We then obtain the location of information transfers by computing the distance between the theme summary and the different utterances produced by a speaker. We evaluate two Large Language Models (LLMs) on this pipeline, on the French conversational corpus Paco-Cheese. More generally, we explore how the recent developments in the field of LLMs provide us with the means to implement these new methods and more generally support research into questions that usually heavily relies on human annotators.", }

Interaction theories suggest that the emergence of mutual understanding between speakers in natural conversations depends on the construction of a shared knowledge base (\textit{common ground}), but the details of which information and the circumstances under which it is memorized are not explained by any model. Previous works have looked at metrics derived from Information Theory to quantify the dynamics of information exchanged between participants, but do not provide an efficient way to locate information that will enter the common ground. We propose a new method based on the segmentation of a conversation into themes followed by their summarization. We then obtain the location of information transfers by computing the distance between the theme summary and the different utterances produced by a speaker. We evaluate two Large Language Models (LLMs) on this pipeline, on the French conversational corpus Paco-Cheese. More generally, we explore how the recent developments in the field of LLMs provide us with the means to implement these new methods and more generally support research into questions that usually heavily relies on human annotators.

Did You Get It? A Zero-Shot Approach to Locate Information Transfers in Conversations

lrec-main.437

Poster

https://aclanthology.org/2024.lrec-main.438.bib

https://aclanthology.org/2024.lrec-main.438/

@inproceedings{lee-etal-2024-difficulty, title = "Difficulty-Focused Contrastive Learning for Knowledge Tracing with a Large Language Model-Based Difficulty Prediction", author = "Lee, Unggi and Yoon, Sungjun and Yun, Joon Seo and Park, Kyoungsoo and Jung, YoungHoon and Stratton, Damji and Kim, Hyeoncheol", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.438", pages = "4891--4900", abstract = "This paper presents novel techniques for enhancing the performance of knowledge tracing (KT) models by focusing on the crucial factor of question and concept difficulty level. Despite the acknowledged significance of difficulty, previous KT research has yet to exploit its potential for model optimization and has struggled to predict difficulty from unseen data. To address these problems, we propose a difficulty-centered contrastive learning method for KT models and a Large Language Model (LLM)-based framework for difficulty prediction. These innovative methods seek to improve the performance of KT models and provide accurate difficulty estimates for unseen data. Our ablation study demonstrates the efficacy of these techniques by demonstrating enhanced KT model performance. Nonetheless, the complex relationship between language and difficulty merits further investigation.", }

This paper presents novel techniques for enhancing the performance of knowledge tracing (KT) models by focusing on the crucial factor of question and concept difficulty level. Despite the acknowledged significance of difficulty, previous KT research has yet to exploit its potential for model optimization and has struggled to predict difficulty from unseen data. To address these problems, we propose a difficulty-centered contrastive learning method for KT models and a Large Language Model (LLM)-based framework for difficulty prediction. These innovative methods seek to improve the performance of KT models and provide accurate difficulty estimates for unseen data. Our ablation study demonstrates the efficacy of these techniques by demonstrating enhanced KT model performance. Nonetheless, the complex relationship between language and difficulty merits further investigation.

Difficulty-Focused Contrastive Learning for Knowledge Tracing with a Large Language Model-Based Difficulty Prediction

lrec-main.438

Poster

2312.11890

https://aclanthology.org/2024.lrec-main.439.bib

https://aclanthology.org/2024.lrec-main.439/

@inproceedings{xin-etal-2024-diffusion, title = "Diffusion Based Counterfactual Augmentation for Dual Sentiment Classification", author = "Xin, Dancheng and Yuan, Jiawei and Li, Yang", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.439", pages = "4901--4911", abstract = "State-of-the-art NLP models have demonstrated exceptional performance across various tasks, including sentiment analysis. However, concerns have been raised about their robustness and susceptibility to systematic biases in both training and test data, which may lead to performance challenges when these models encounter out-of-distribution data in real-world applications. Although various data augmentation and adversarial perturbation techniques have shown promise in tackling these issues, prior methods such as word embedding perturbation or synonymous sentence expansion have failed to mitigate the spurious association problem inherent in the original data. Recent counterfactual augmentation methods have attempted to tackle this issue, but they have been limited by rigid rules, resulting in inconsistent context and disrupted semantics. In response to these challenges, we introduce a diffusion-based counterfactual data augmentation (DCA) framework. It utilizes an antonymous paradigm to guide the continuous diffusion model and employs reinforcement learning in combination with contrastive learning to optimize algorithms for generating counterfactual samples with high diversity and quality. Furthermore, we use a dual sentiment classifier to validate the generated antonymous samples and subsequently perform sentiment classification. Our experiments on four benchmark datasets demonstrate that DCA achieves state-of-the-art performance in sentiment classification tasks.", }

State-of-the-art NLP models have demonstrated exceptional performance across various tasks, including sentiment analysis. However, concerns have been raised about their robustness and susceptibility to systematic biases in both training and test data, which may lead to performance challenges when these models encounter out-of-distribution data in real-world applications. Although various data augmentation and adversarial perturbation techniques have shown promise in tackling these issues, prior methods such as word embedding perturbation or synonymous sentence expansion have failed to mitigate the spurious association problem inherent in the original data. Recent counterfactual augmentation methods have attempted to tackle this issue, but they have been limited by rigid rules, resulting in inconsistent context and disrupted semantics. In response to these challenges, we introduce a diffusion-based counterfactual data augmentation (DCA) framework. It utilizes an antonymous paradigm to guide the continuous diffusion model and employs reinforcement learning in combination with contrastive learning to optimize algorithms for generating counterfactual samples with high diversity and quality. Furthermore, we use a dual sentiment classifier to validate the generated antonymous samples and subsequently perform sentiment classification. Our experiments on four benchmark datasets demonstrate that DCA achieves state-of-the-art performance in sentiment classification tasks.

Diffusion Based Counterfactual Augmentation for Dual Sentiment Classification

lrec-main.439

Poster

https://aclanthology.org/2024.lrec-main.440.bib

https://aclanthology.org/2024.lrec-main.440/

@inproceedings{xiang-etal-2024-diffusiondialog, title = "{D}iffusion{D}ialog: A Diffusion Model for Diverse Dialog Generation with Latent Space", author = "Xiang, Jianxiang and Liu, Zhenhua and Liu, Haodong and Bai, Yin and Cheng, Jia and Chen, Wenliang", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.440", pages = "4912--4921", abstract = "In real-life conversations, the content is diverse, and there exist one-to-many problems that require diverse generation. Previous studies attempted to introduce discrete or Gaussian-based latent variables to address the one-to-many problem, but the diversity is limited. Recently, diffusion models have made breakthroughs in computer vision and some attempts have been made in natural language processing. In this paper, we propose DiffusionDialog, a novel approach to enhance the diversity of dialogue generation with the help of diffusion model. In our approach, we introduce the continuous latent variables in the diffusion model instead of the discrete ones or VAE, which are often used in the previous studies. The problem of using discrete variables in dialog task is how to build a effective prior of latent space and inferring process to infer the proper latent given the context. Combining the encoder and latent-based diffusion model, we encode the latent of response in a continuous space as the prior instead of fixed Gaussian distribution in VAE or simply discrete ones, and we infer the latent by denoising step by step with diffusion model. The experimental results show that our model greatly enhance the diversity of dialog response while keeping the coherence. In further analysis, we find that our diffusion model achieved high inference efficiency which is the main challenge of applying diffusion model in natural language processing.", }

In real-life conversations, the content is diverse, and there exist one-to-many problems that require diverse generation. Previous studies attempted to introduce discrete or Gaussian-based latent variables to address the one-to-many problem, but the diversity is limited. Recently, diffusion models have made breakthroughs in computer vision and some attempts have been made in natural language processing. In this paper, we propose DiffusionDialog, a novel approach to enhance the diversity of dialogue generation with the help of diffusion model. In our approach, we introduce the continuous latent variables in the diffusion model instead of the discrete ones or VAE, which are often used in the previous studies. The problem of using discrete variables in dialog task is how to build a effective prior of latent space and inferring process to infer the proper latent given the context. Combining the encoder and latent-based diffusion model, we encode the latent of response in a continuous space as the prior instead of fixed Gaussian distribution in VAE or simply discrete ones, and we infer the latent by denoising step by step with diffusion model. The experimental results show that our model greatly enhance the diversity of dialog response while keeping the coherence. In further analysis, we find that our diffusion model achieved high inference efficiency which is the main challenge of applying diffusion model in natural language processing.

DiffusionDialog: A Diffusion Model for Diverse Dialog Generation with Latent Space

lrec-main.440

Poster

2404.06760

https://aclanthology.org/2024.lrec-main.441.bib

https://aclanthology.org/2024.lrec-main.441/

@inproceedings{zhang-etal-2024-dima, title = "{D}im{A}: A Parameter-efficient Fine-tuning Method with Knowledge Transfer Based on Transformer", author = "Zhang, Wenxuan and Huang, Min and Song, Zhuoyang and Miao, Qinghai", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.441", pages = "4922--4934", abstract = "Fine-tuning is a widely used technique for leveraging pre-trained language models (PLMs) in downstream tasks, but it can be computationally expensive and storage-intensive. To address this challenge, researchers have developed parameter-efficient methods that balance performance and resource cost. However, these methods often come with trade-offs like increased inference latency, token length usage, or limited adaptability for multitasking scenarios. This paper introduces a novel parameter-efficient method called DimA(Dimensionality Augmentation), which enhances the Transformer architecture by increasing the dimensionality. DimA achieves state-of-the-art results in GLUE and XSUM tasks while utilizing less than 1{\%} of the original model{'}s parameters. Moreover, DimA introduces a novel approach to knowledge transfer that enables the simultaneous utilization of knowledge learned from multiple tasks to handle new tasks. This method significantly enhances the performance of the model on new tasks. Its versatility in model structure also enables its application to various Transformer-based models.", }

Fine-tuning is a widely used technique for leveraging pre-trained language models (PLMs) in downstream tasks, but it can be computationally expensive and storage-intensive. To address this challenge, researchers have developed parameter-efficient methods that balance performance and resource cost. However, these methods often come with trade-offs like increased inference latency, token length usage, or limited adaptability for multitasking scenarios. This paper introduces a novel parameter-efficient method called DimA(Dimensionality Augmentation), which enhances the Transformer architecture by increasing the dimensionality. DimA achieves state-of-the-art results in GLUE and XSUM tasks while utilizing less than 1{\%} of the original model{'}s parameters. Moreover, DimA introduces a novel approach to knowledge transfer that enables the simultaneous utilization of knowledge learned from multiple tasks to handle new tasks. This method significantly enhances the performance of the model on new tasks. Its versatility in model structure also enables its application to various Transformer-based models.

DimA: A Parameter-efficient Fine-tuning Method with Knowledge Transfer Based on Transformer

lrec-main.441

Poster

https://aclanthology.org/2024.lrec-main.442.bib

https://aclanthology.org/2024.lrec-main.442/

@inproceedings{sato-2024-disambiguating, title = "Disambiguating Homographs and Homophones Simultaneously: A Regrouping Method for {J}apanese", author = "Sato, Yo", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.442", pages = "4935--4939", abstract = "We present a method that re-groups surface forms into clusters representing synonyms, and help disambiguate homographs as well as homophone. The method is applied post-hoc to trained contextual word embeddings. It is beneficial to languages where both homographs and homophones abound, which compromise the efficiency of language model and causes the underestimation problem in evaluation. Taking Japanese as an example, we evaluate how accurate such disambiguation can be, and how much the underestimation can be mitigated.", }

We present a method that re-groups surface forms into clusters representing synonyms, and help disambiguate homographs as well as homophone. The method is applied post-hoc to trained contextual word embeddings. It is beneficial to languages where both homographs and homophones abound, which compromise the efficiency of language model and causes the underestimation problem in evaluation. Taking Japanese as an example, we evaluate how accurate such disambiguation can be, and how much the underestimation can be mitigated.

Disambiguating Homographs and Homophones Simultaneously: A Regrouping Method for Japanese

lrec-main.442

Poster

https://aclanthology.org/2024.lrec-main.443.bib

https://aclanthology.org/2024.lrec-main.443/

@inproceedings{yung-etal-2024-discogem, title = "{D}isco{G}e{M} 2.0: A Parallel Corpus of {E}nglish, {G}erman, {F}rench and {C}zech Implicit Discourse Relations", author = "Yung, Frances and Scholman, Merel and Zikanova, Sarka and Demberg, Vera", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.443", pages = "4940--4956", abstract = "We present DiscoGeM 2.0, a crowdsourced, parallel corpus of 12,834 implicit discourse relations, with English, German, French and Czech data. We propose and validate a new single-step crowdsourcing annotation method and apply it to collect new annotations in German, French and Czech. The corpus was constructed by having crowdsourced annotators choose a suitable discourse connective for each relation from a set of unambiguous candidates. Every instance was annotated by 10 workers. Our corpus hence represents the first multi-lingual resource that contains distributions of discourse interpretations for implicit relations. The results show that the connective insertion method of discourse annotation can be reliably extended to other languages. The resulting multi-lingual annotations also reveal that implicit relations inferred in one language may differ from those inferred in the translation, meaning the annotations are not always directly transferable. DiscoGem 2.0 promotes the investigation of cross-linguistic differences in discourse marking and could improve automatic discourse parsing applications. It is openly downloadable here: https://github.com/merelscholman/DiscoGeM.", }

We present DiscoGeM 2.0, a crowdsourced, parallel corpus of 12,834 implicit discourse relations, with English, German, French and Czech data. We propose and validate a new single-step crowdsourcing annotation method and apply it to collect new annotations in German, French and Czech. The corpus was constructed by having crowdsourced annotators choose a suitable discourse connective for each relation from a set of unambiguous candidates. Every instance was annotated by 10 workers. Our corpus hence represents the first multi-lingual resource that contains distributions of discourse interpretations for implicit relations. The results show that the connective insertion method of discourse annotation can be reliably extended to other languages. The resulting multi-lingual annotations also reveal that implicit relations inferred in one language may differ from those inferred in the translation, meaning the annotations are not always directly transferable. DiscoGem 2.0 promotes the investigation of cross-linguistic differences in discourse marking and could improve automatic discourse parsing applications. It is openly downloadable here: https://github.com/merelscholman/DiscoGeM.

DiscoGeM 2.0: A Parallel Corpus of English, German, French and Czech Implicit Discourse Relations

lrec-main.443

Poster

https://aclanthology.org/2024.lrec-main.444.bib

https://aclanthology.org/2024.lrec-main.444/

@inproceedings{thompson-etal-2024-discourse, title = "Discourse Structure for the {M}inecraft Corpus", author = "Thompson, Kate and Hunter, Julie and Asher, Nicholas", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.444", pages = "4957--4967", abstract = "We provide a new linguistic resource: The Minecraft Structured Dialogue Corpus (MSDC), a discourse annotated version of the Minecraft Dialogue Corpus (MDC; Narayan-Chen et al., 2019), with complete, situated discourse structures in the style of SDRT (Asher and Lascarides, 2003). Our structures feature both linguistic discourse moves and nonlinguistic actions. To show computational tractability, we train a discourse parser with a novel {``}2 pass architecture{''} on MSDC that gives excellent results on attachment prediction and relation labeling tasks especially long distance attachments.", }

We provide a new linguistic resource: The Minecraft Structured Dialogue Corpus (MSDC), a discourse annotated version of the Minecraft Dialogue Corpus (MDC; Narayan-Chen et al., 2019), with complete, situated discourse structures in the style of SDRT (Asher and Lascarides, 2003). Our structures feature both linguistic discourse moves and nonlinguistic actions. To show computational tractability, we train a discourse parser with a novel {``}2 pass architecture{''} on MSDC that gives excellent results on attachment prediction and relation labeling tasks especially long distance attachments.

Discourse Structure for the Minecraft Corpus

lrec-main.444

Poster

https://aclanthology.org/2024.lrec-main.445.bib

https://aclanthology.org/2024.lrec-main.445/

@inproceedings{xie-li-2024-discriminative, title = "Discriminative Language Model as Semantic Consistency Scorer for Prompt-based Few-Shot Text Classification", author = "Xie, Zhipeng and Li, Yahe", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.445", pages = "4968--4977", abstract = "A successful prompt-based finetuning method should have three prerequisites: task compatibility, input compatibility, and evidence abundance. Bearing this belief in mind, this paper designs a novel prompt-based method (called DLM-SCS) for few-shot text classification, which utilizes the discriminative language model ELECTRA that is pretrained to distinguish whether a token is original or replaced. The method is built upon the intuitive idea that the prompt instantiated with the true label should have higher semantic consistency score than other prompts with false labels. Since a prompt usually consists of several components (or parts), its semantic consistency can be decomposed accordingly, which means each part can provide information for semantic consistency discrimination. The semantic consistency of each component is then computed by making use of the pretrained ELECTRA model, where no extra parameters get introduced. Extensive experiments have shown that our model outperforms several state-of-the-art prompt-based few-shot methods on 10 widely-used text classification tasks.", }

A successful prompt-based finetuning method should have three prerequisites: task compatibility, input compatibility, and evidence abundance. Bearing this belief in mind, this paper designs a novel prompt-based method (called DLM-SCS) for few-shot text classification, which utilizes the discriminative language model ELECTRA that is pretrained to distinguish whether a token is original or replaced. The method is built upon the intuitive idea that the prompt instantiated with the true label should have higher semantic consistency score than other prompts with false labels. Since a prompt usually consists of several components (or parts), its semantic consistency can be decomposed accordingly, which means each part can provide information for semantic consistency discrimination. The semantic consistency of each component is then computed by making use of the pretrained ELECTRA model, where no extra parameters get introduced. Extensive experiments have shown that our model outperforms several state-of-the-art prompt-based few-shot methods on 10 widely-used text classification tasks.

Discriminative Language Model as Semantic Consistency Scorer for Prompt-based Few-Shot Text Classification

lrec-main.445

Poster

2210.12763

https://aclanthology.org/2024.lrec-main.446.bib

https://aclanthology.org/2024.lrec-main.446/

@inproceedings{hudi-etal-2024-disentangling, title = "Disentangling Pretrained Representation to Leverage Low-Resource Languages in Multilingual Machine Translation", author = "Hudi, Frederikus and Qu, Zhi and Kamigaito, Hidetaka and Watanabe, Taro", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.446", pages = "4978--4989", abstract = "Multilingual neural machine translation aims to encapsulate multiple languages into a single model. However, it requires an enormous dataset, leaving the low-resource language (LRL) underdeveloped. As LRLs may benefit from shared knowledge of multilingual representation, we aspire to find effective ways to integrate unseen languages in a pre-trained model. Nevertheless, the intricacy of shared representation among languages hinders its full utilisation. To resolve this problem, we employed target language prediction and a central language-aware layer to improve representation in integrating LRLs. Focusing on improving LRLs in the linguistically diverse country of Indonesia, we evaluated five languages using a parallel corpus of 1,000 instances each, with experimental results measured by BLEU showing zero-shot improvement of 7.4 from the baseline score of 7.1 to a score of 15.5 at best. Further analysis showed that the gains in performance are attributed more to the disentanglement of multilingual representation in the encoder with the shift of the target language-specific representation in the decoder.", }

Multilingual neural machine translation aims to encapsulate multiple languages into a single model. However, it requires an enormous dataset, leaving the low-resource language (LRL) underdeveloped. As LRLs may benefit from shared knowledge of multilingual representation, we aspire to find effective ways to integrate unseen languages in a pre-trained model. Nevertheless, the intricacy of shared representation among languages hinders its full utilisation. To resolve this problem, we employed target language prediction and a central language-aware layer to improve representation in integrating LRLs. Focusing on improving LRLs in the linguistically diverse country of Indonesia, we evaluated five languages using a parallel corpus of 1,000 instances each, with experimental results measured by BLEU showing zero-shot improvement of 7.4 from the baseline score of 7.1 to a score of 15.5 at best. Further analysis showed that the gains in performance are attributed more to the disentanglement of multilingual representation in the encoder with the shift of the target language-specific representation in the decoder.

Disentangling Pretrained Representation to Leverage Low-Resource Languages in Multilingual Machine Translation

lrec-main.446

Poster

https://aclanthology.org/2024.lrec-main.447.bib

https://aclanthology.org/2024.lrec-main.447/

@inproceedings{braud-etal-2024-disrpt, title = "{DISRPT}: A Multilingual, Multi-domain, Cross-framework Benchmark for Discourse Processing", author = "Braud, Chlo{\'e} and Zeldes, Amir and Rivi{\`e}re, Laura and Liu, Yang Janet and Muller, Philippe and Sileo, Damien and Aoyama, Tatsuya", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.447", pages = "4990--5005", abstract = "This paper presents DISRPT, a multilingual, multi-domain, and cross-framework benchmark dataset for discourse processing, covering the tasks of discourse unit segmentation, connective identification, and relation classification. DISRPT includes 13 languages, with data from 24 corpora covering about 4 millions tokens and around 250,000 discourse relation instances from 4 discourse frameworks: RST, SDRT, PDTB, and Discourse Dependencies. We present an overview of the data, its development across three NLP shared tasks on discourse processing carried out in the past five years, and the latest modifications and added extensions. We also carry out an evaluation of state-of-the-art multilingual systems trained on the data for each task, showing plateau performance on segmentation, but important room for improvement for connective identification and relation classification. The DISRPT benchmark employs a unified format that we make available on GitHub and HuggingFace in order to encourage future work on discourse processing across languages, domains, and frameworks.", }

This paper presents DISRPT, a multilingual, multi-domain, and cross-framework benchmark dataset for discourse processing, covering the tasks of discourse unit segmentation, connective identification, and relation classification. DISRPT includes 13 languages, with data from 24 corpora covering about 4 millions tokens and around 250,000 discourse relation instances from 4 discourse frameworks: RST, SDRT, PDTB, and Discourse Dependencies. We present an overview of the data, its development across three NLP shared tasks on discourse processing carried out in the past five years, and the latest modifications and added extensions. We also carry out an evaluation of state-of-the-art multilingual systems trained on the data for each task, showing plateau performance on segmentation, but important room for improvement for connective identification and relation classification. The DISRPT benchmark employs a unified format that we make available on GitHub and HuggingFace in order to encourage future work on discourse processing across languages, domains, and frameworks.

DISRPT: A Multilingual, Multi-domain, Cross-framework Benchmark for Discourse Processing

lrec-main.447

Poster

https://aclanthology.org/2024.lrec-main.448.bib

https://aclanthology.org/2024.lrec-main.448/

@inproceedings{liang-etal-2024-distantly, title = "Distantly Supervised Contrastive Learning for Low-Resource Scripting Language Summarization", author = "Liang, Junzhe and Sun, Haifeng and Zhuang, Zirui and Qi, Qi and Wang, Jingyu and Liao, Jianxin", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.448", pages = "5006--5017", abstract = "Code summarization provides a natural language description for a given piece of code. In this work, we focus on scripting code{---}programming languages that interact with specific devices through commands. The low-resource nature of scripting languages makes traditional code summarization methods challenging to apply. To address this, we introduce a novel framework: distantly supervised contrastive learning for low-resource scripting language summarization. This framework leverages limited atomic commands and category constraints to enhance code representations. Extensive experiments demonstrate our method{'}s superiority over competitive baselines.", }

Code summarization provides a natural language description for a given piece of code. In this work, we focus on scripting code{---}programming languages that interact with specific devices through commands. The low-resource nature of scripting languages makes traditional code summarization methods challenging to apply. To address this, we introduce a novel framework: distantly supervised contrastive learning for low-resource scripting language summarization. This framework leverages limited atomic commands and category constraints to enhance code representations. Extensive experiments demonstrate our method{'}s superiority over competitive baselines.

Distantly Supervised Contrastive Learning for Low-Resource Scripting Language Summarization

lrec-main.448

Poster

https://aclanthology.org/2024.lrec-main.449.bib

https://aclanthology.org/2024.lrec-main.449/

@inproceedings{zhang-etal-2024-distillation, title = "Distillation with Explanations from Large Language Models", author = "Zhang, Hanyu and Wang, Xiting and Ao, Xiang and He, Qing", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.449", pages = "5018--5028", abstract = "Free-text explanations are crucial for enhancing the interpretability of AI models. However, training models to generate high-quality free-text explanations is challenging, primarily due to the requirement of a substantial amount of human-written explanations, which can be expensive. Recently, Large language models (LLMs) like ChatGPT and GPT-4 have made remarkable progress in various NLP tasks while also providing explanations alongside their answers. Leveraging LLMs for data labeling offers a more cost-effective alternative. However, a key concern arises from the fact that the answers provided by LLMs are not entirely accurate, potentially introducing noise to both task outputs and explanation generation. To remedy this, we propose a new mechanism, Distillation with Explanations from LLMs. we observe that despite the incorrectness in LLMs-generated answers, their explanations are consistent with their answers. Leveraging this consistency, our method combines the ground truth labels and answers-explanations generated by LLMs, to simultaneously generate more accurate answers and the corresponding free-text explanations. Experimental results demonstrate that our approach achieves improved predictive performance and also generates explanations that exhibit greater alignment with the model{'}s task outputs.", }

Free-text explanations are crucial for enhancing the interpretability of AI models. However, training models to generate high-quality free-text explanations is challenging, primarily due to the requirement of a substantial amount of human-written explanations, which can be expensive. Recently, Large language models (LLMs) like ChatGPT and GPT-4 have made remarkable progress in various NLP tasks while also providing explanations alongside their answers. Leveraging LLMs for data labeling offers a more cost-effective alternative. However, a key concern arises from the fact that the answers provided by LLMs are not entirely accurate, potentially introducing noise to both task outputs and explanation generation. To remedy this, we propose a new mechanism, Distillation with Explanations from LLMs. we observe that despite the incorrectness in LLMs-generated answers, their explanations are consistent with their answers. Leveraging this consistency, our method combines the ground truth labels and answers-explanations generated by LLMs, to simultaneously generate more accurate answers and the corresponding free-text explanations. Experimental results demonstrate that our approach achieves improved predictive performance and also generates explanations that exhibit greater alignment with the model{'}s task outputs.

Distillation with Explanations from Large Language Models

lrec-main.449

Poster

https://aclanthology.org/2024.lrec-main.450.bib

https://aclanthology.org/2024.lrec-main.450/

@inproceedings{huang-etal-2024-distill, title = "Distill, Fuse, Pre-train: Towards Effective Event Causality Identification with Commonsense-Aware Pre-trained Model", author = "Huang, Peixin and Zhao, Xiang and Hu, Minghao and Tan, Zhen and Xiao, Weidong", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.450", pages = "5029--5040", abstract = "Event Causality Identification (ECI) aims to detect causal relations between events in unstructured texts. This task is challenged by the lack of data and explicit causal clues. Some methods incorporate explicit knowledge from external knowledge graphs (KGs) into Pre-trained Language Models (PLMs) to tackle these issues, achieving certain accomplishments. However, they ignore that existing KGs usually contain trivial knowledge which may prejudice the performance. Moreover, they simply integrate the concept triplets, underutilizing the deep interaction between the text and external graph. In this paper, we propose an effective pipeline DFP, i.e., Distill, Fuse and Pre-train, to build a commonsense-aware pre-trained model which integrates reliable task-specific knowledge from commonsense graphs. This pipeline works as follows: (1) To leverage the reliable knowledge, commonsense graph distillation is proposed to distill commonsense graphs and obtain the meta-graph which contain credible task-oriented knowledge. (2) To model the deep interaction between the text and external graph, heterogeneous information fusion is proposed to fuse them through a commonsense-aware memory network. (3) Continual pre-training designs three continual pre-training tasks to further align and fuse the text and the commonsense meta-graph. Through extensive experiments on two benchmarks, we demonstrate the validity of our pipeline.", }

Event Causality Identification (ECI) aims to detect causal relations between events in unstructured texts. This task is challenged by the lack of data and explicit causal clues. Some methods incorporate explicit knowledge from external knowledge graphs (KGs) into Pre-trained Language Models (PLMs) to tackle these issues, achieving certain accomplishments. However, they ignore that existing KGs usually contain trivial knowledge which may prejudice the performance. Moreover, they simply integrate the concept triplets, underutilizing the deep interaction between the text and external graph. In this paper, we propose an effective pipeline DFP, i.e., Distill, Fuse and Pre-train, to build a commonsense-aware pre-trained model which integrates reliable task-specific knowledge from commonsense graphs. This pipeline works as follows: (1) To leverage the reliable knowledge, commonsense graph distillation is proposed to distill commonsense graphs and obtain the meta-graph which contain credible task-oriented knowledge. (2) To model the deep interaction between the text and external graph, heterogeneous information fusion is proposed to fuse them through a commonsense-aware memory network. (3) Continual pre-training designs three continual pre-training tasks to further align and fuse the text and the commonsense meta-graph. Through extensive experiments on two benchmarks, we demonstrate the validity of our pipeline.

Distill, Fuse, Pre-train: Towards Effective Event Causality Identification with Commonsense-Aware Pre-trained Model

lrec-main.450

Poster

https://aclanthology.org/2024.lrec-main.451.bib

https://aclanthology.org/2024.lrec-main.451/

@inproceedings{ye-etal-2024-distilling, title = "Distilling Causal Effect of Data in Continual Few-shot Relation Learning", author = "Ye, Weihang and Zhang, Peng and Zhang, Jing and Gao, Hui and Wang, Moyao", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.451", pages = "5041--5051", abstract = "Continual Few-Shot Relation Learning (CFRL) aims to learn an increasing number of new relational patterns from a data stream. However, due to the limited number of samples and the continual training mode, this method frequently encounters the catastrophic forgetting issues. The research on causal inference suggests that this issue is caused by the loss of causal effects from old data during the new training process. Inspired by the causal graph, we propose a unified causal framework for CFRL to restore the causal effects. Specifically, we establish two additional causal paths from old data to predictions by having the new data and memory data collide with old data separately in the old feature space. This augmentation allows us to preserve causal effects effectively and enhance the utilization of valuable information within memory data, thereby alleviating the phenomenon of catastrophic forgetting. Furthermore, we introduce a self-adaptive weight to achieve a delicate balance of causal effects between the new and old relation types. Extensive experiments demonstrate the superiority of our method over existing state-of-the-art approaches in CFRL task settings. Our codes are publicly available at: https://github.com/ywh140/CECF.", }

Continual Few-Shot Relation Learning (CFRL) aims to learn an increasing number of new relational patterns from a data stream. However, due to the limited number of samples and the continual training mode, this method frequently encounters the catastrophic forgetting issues. The research on causal inference suggests that this issue is caused by the loss of causal effects from old data during the new training process. Inspired by the causal graph, we propose a unified causal framework for CFRL to restore the causal effects. Specifically, we establish two additional causal paths from old data to predictions by having the new data and memory data collide with old data separately in the old feature space. This augmentation allows us to preserve causal effects effectively and enhance the utilization of valuable information within memory data, thereby alleviating the phenomenon of catastrophic forgetting. Furthermore, we introduce a self-adaptive weight to achieve a delicate balance of causal effects between the new and old relation types. Extensive experiments demonstrate the superiority of our method over existing state-of-the-art approaches in CFRL task settings. Our codes are publicly available at: https://github.com/ywh140/CECF.

Distilling Causal Effect of Data in Continual Few-shot Relation Learning

lrec-main.451

Poster

https://aclanthology.org/2024.lrec-main.452.bib

https://aclanthology.org/2024.lrec-main.452/

@inproceedings{taslimipoor-etal-2024-distractor, title = "Distractor Generation Using Generative and Discriminative Capabilities of Transformer-based Models", author = "Taslimipoor, Shiva and Benedetto, Luca and Felice, Mariano and Buttery, Paula", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.452", pages = "5052--5063", abstract = "Multiple Choice Questions (MCQs) are very common in both high-stakes and low-stakes examinations, and their effectiveness in assessing students relies on the quality and diversity of distractors, which are the incorrect answer options provided alongside the correct answer. Motivated by the progress in generative language models, we propose a two-step automatic distractor generation approach which is based on text to text transfer transformer models. Unlike most of previous methods for distractor generation, our approach does not rely on the correct answer options. Instead, it first generates both correct and incorrect answer options, and then discriminates potential correct options from distractors. Identified distractors are finally categorised based on semantic similarity scores into separate clusters, and the cluster heads are selected as our final distinct distractors. Experiments on two publicly available datasets show that our approach outperforms previous models both in the case of single-word answer options and longer-sequence reading comprehension questions.", }

Multiple Choice Questions (MCQs) are very common in both high-stakes and low-stakes examinations, and their effectiveness in assessing students relies on the quality and diversity of distractors, which are the incorrect answer options provided alongside the correct answer. Motivated by the progress in generative language models, we propose a two-step automatic distractor generation approach which is based on text to text transfer transformer models. Unlike most of previous methods for distractor generation, our approach does not rely on the correct answer options. Instead, it first generates both correct and incorrect answer options, and then discriminates potential correct options from distractors. Identified distractors are finally categorised based on semantic similarity scores into separate clusters, and the cluster heads are selected as our final distinct distractors. Experiments on two publicly available datasets show that our approach outperforms previous models both in the case of single-word answer options and longer-sequence reading comprehension questions.

Distractor Generation Using Generative and Discriminative Capabilities of Transformer-based Models

lrec-main.452

Poster

https://aclanthology.org/2024.lrec-main.453.bib

https://aclanthology.org/2024.lrec-main.453/

@inproceedings{chan-etal-2024-distribution, title = "Distribution Aware Metrics for Conditional Natural Language Generation", author = "Chan, David M. and Ni, Yiming and Ross, David and Vijayanarasimhan, Sudheendra and Myers, Austin and Canny, John", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.453", pages = "5064--5095", abstract = "Traditional automated metrics for evaluating conditional natural language generation rely on pairwise comparisons between a single generated text and the best-matching gold-standard reference. This method is effective when ground truth data diversity can be attributed to noise, however, it falls short when diversity in references holds valuable contextual information, as in visual description or summarization, as it does not evaluate the ability of a model to generate text matching the diversity of the ground truth samples. In this paper, we challenge the adequacy of existing metrics in such semantically diverse contexts and introduce a novel approach for evaluating conditional language generation models, leveraging a family of meta-metrics that build on existing pairwise distance functions. These meta-metrics assess not just single-samples, but distributions of reference and model-generated captions using small sample sets. We demonstrate our approach through a case study of visual description in the English language which reveals not only how current models prioritize single-description quality over diversity, but further sheds light on the impact of sampling methods and temperature settings on description quality and diversity.", }

Traditional automated metrics for evaluating conditional natural language generation rely on pairwise comparisons between a single generated text and the best-matching gold-standard reference. This method is effective when ground truth data diversity can be attributed to noise, however, it falls short when diversity in references holds valuable contextual information, as in visual description or summarization, as it does not evaluate the ability of a model to generate text matching the diversity of the ground truth samples. In this paper, we challenge the adequacy of existing metrics in such semantically diverse contexts and introduce a novel approach for evaluating conditional language generation models, leveraging a family of meta-metrics that build on existing pairwise distance functions. These meta-metrics assess not just single-samples, but distributions of reference and model-generated captions using small sample sets. We demonstrate our approach through a case study of visual description in the English language which reveals not only how current models prioritize single-description quality over diversity, but further sheds light on the impact of sampling methods and temperature settings on description quality and diversity.

Distribution Aware Metrics for Conditional Natural Language Generation

lrec-main.453

Poster

2209.07518

https://aclanthology.org/2024.lrec-main.454.bib

https://aclanthology.org/2024.lrec-main.454/

@inproceedings{guo-etal-2024-diversifying, title = "Diversifying Question Generation over Knowledge Base via External Natural Questions", author = "Guo, Shasha and Zhang, Jing and Ke, Xirui and Li, Cuiping and Chen, Hong", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.454", pages = "5096--5108", abstract = "Previous methods on knowledge base question generation (KBQG) primarily focus on refining the quality of a single generated question. However, considering the remarkable paraphrasing ability of humans, we believe that diverse texts can express identical semantics through varied expressions. The above insights make diversifying question generation an intriguing task, where the first challenge is evaluation metrics for diversity. Current metrics inadequately assess the aforementioned diversity. They calculate the ratio of unique n-grams in the generated question, which tends to measure duplication rather than true diversity. Accordingly, we devise a new diversity evaluation metric, which measures the diversity among top-k generated questions for each instance while ensuring their relevance to the ground truth. Clearly, the second challenge is how to enhance diversifying question generation. To address this challenge, we introduce a dual model framework interwoven by two selection strategies to generate diverse questions leveraging external natural questions. The main idea of our dual framework is to extract more diverse expressions and integrate them into the generation model to enhance diversifying question generation. Extensive experiments on widely used benchmarks for KBQG show that our approach can outperform pre-trained language model baselines and text-davinci-003 in diversity while achieving comparable performance with ChatGPT.", }

Previous methods on knowledge base question generation (KBQG) primarily focus on refining the quality of a single generated question. However, considering the remarkable paraphrasing ability of humans, we believe that diverse texts can express identical semantics through varied expressions. The above insights make diversifying question generation an intriguing task, where the first challenge is evaluation metrics for diversity. Current metrics inadequately assess the aforementioned diversity. They calculate the ratio of unique n-grams in the generated question, which tends to measure duplication rather than true diversity. Accordingly, we devise a new diversity evaluation metric, which measures the diversity among top-k generated questions for each instance while ensuring their relevance to the ground truth. Clearly, the second challenge is how to enhance diversifying question generation. To address this challenge, we introduce a dual model framework interwoven by two selection strategies to generate diverse questions leveraging external natural questions. The main idea of our dual framework is to extract more diverse expressions and integrate them into the generation model to enhance diversifying question generation. Extensive experiments on widely used benchmarks for KBQG show that our approach can outperform pre-trained language model baselines and text-davinci-003 in diversity while achieving comparable performance with ChatGPT.

Diversifying Question Generation over Knowledge Base via External Natural Questions

lrec-main.454

Poster

2309.14362

https://aclanthology.org/2024.lrec-main.455.bib

https://aclanthology.org/2024.lrec-main.455/

@inproceedings{wei-etal-2024-dmon, title = "{DMON}: A Simple Yet Effective Approach for Argument Structure Learning", author = "Wei, Sun and Li, Mingxiao and Sun, Jingyuan and Davis, Jesse and Moens, Marie-Francine", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.455", pages = "5109--5118", abstract = "Argument structure learning (ASL) entails predicting relations between arguments. Because it can structure a document to facilitate its understanding, it has been widely applied in many fields (medical, commercial, and scientific domains). Despite its broad utilization, ASL remains a challenging task because it involves examining the complex relationships between the sentences in a potentially unstructured discourse. To resolve this problem, we have developed a simple yet effective approach called Dual-tower Multi-scale cOnvolution neural Network (DMON) for the ASL task. Specifically, we organize arguments into a relationship matrix that together with the argument embeddings forms a relationship tensor and design a mechanism to capture relations with contextual arguments. Experimental results on three different-domain argument mining datasets demonstrate that our framework outperforms state-of-the-art models. We will release the code after paper acceptance.", }

Argument structure learning (ASL) entails predicting relations between arguments. Because it can structure a document to facilitate its understanding, it has been widely applied in many fields (medical, commercial, and scientific domains). Despite its broad utilization, ASL remains a challenging task because it involves examining the complex relationships between the sentences in a potentially unstructured discourse. To resolve this problem, we have developed a simple yet effective approach called Dual-tower Multi-scale cOnvolution neural Network (DMON) for the ASL task. Specifically, we organize arguments into a relationship matrix that together with the argument embeddings forms a relationship tensor and design a mechanism to capture relations with contextual arguments. Experimental results on three different-domain argument mining datasets demonstrate that our framework outperforms state-of-the-art models. We will release the code after paper acceptance.

DMON: A Simple Yet Effective Approach for Argument Structure Learning

lrec-main.455

Poster

2405.01216

https://aclanthology.org/2024.lrec-main.456.bib

https://aclanthology.org/2024.lrec-main.456/

@inproceedings{zhu-etal-2024-doc2soargraph, title = "{D}oc2{S}oar{G}raph: Discrete Reasoning over Visually-Rich Table-Text Documents via Semantic-Oriented Hierarchical Graphs", author = "Zhu, Fengbin and Wang, Chao and Feng, Fuli and Ren, Zifeng and Li, Moxin and Chua, Tat-Seng", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.456", pages = "5119--5131", abstract = "Table-text document (e.g., financial reports) understanding has attracted increasing attention in recent two years. TAT-DQA is a realistic setting for the understanding of visually-rich table-text documents, which involves answering associated questions requiring discrete reasoning. Most existing work relies on token-level semantics, falling short in the reasoning across document elements such as quantities and dates. To address this limitation, we propose a novel Doc2SoarGraph model that exploits element-level semantics and employs Semantic-oriented hierarchical Graph structures to capture the differences and correlations among different elements within the given document and question. Extensive experiments on the TAT-DQA dataset reveal that our model surpasses the state-of-the-art conventional method (i.e., MHST) and large language model (i.e., ChatGPT) by 17.73 and 6.49 points respectively in terms of Exact Match (EM) metric, demonstrating exceptional effectiveness.", }

Table-text document (e.g., financial reports) understanding has attracted increasing attention in recent two years. TAT-DQA is a realistic setting for the understanding of visually-rich table-text documents, which involves answering associated questions requiring discrete reasoning. Most existing work relies on token-level semantics, falling short in the reasoning across document elements such as quantities and dates. To address this limitation, we propose a novel Doc2SoarGraph model that exploits element-level semantics and employs Semantic-oriented hierarchical Graph structures to capture the differences and correlations among different elements within the given document and question. Extensive experiments on the TAT-DQA dataset reveal that our model surpasses the state-of-the-art conventional method (i.e., MHST) and large language model (i.e., ChatGPT) by 17.73 and 6.49 points respectively in terms of Exact Match (EM) metric, demonstrating exceptional effectiveness.

Doc2SoarGraph: Discrete Reasoning over Visually-Rich Table-Text Documents via Semantic-Oriented Hierarchical Graphs

lrec-main.456

Poster

2305.01938

https://aclanthology.org/2024.lrec-main.457.bib

https://aclanthology.org/2024.lrec-main.457/

@inproceedings{mathur-etal-2024-doc, title = "{DOC}-{RAG}: {ASR} Language Model Personalization with Domain-Distributed Co-occurrence Retrieval Augmentation", author = "Mathur, Puneet and Liu, Zhe and Li, Ke and Ma, Yingyi and Karen, Gil and Ahmed, Zeeshan and Manocha, Dinesh and Zhang, Xuedong", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.457", pages = "5132--5139", abstract = "We propose DOC-RAG - Domain-distributed Co-occurrence Retrieval Augmentation for ASR language model personalization aiming to improve the automatic speech recognition of rare word patterns in unseen domains. Our approach involves contrastively training a document retrieval module to rank external knowledge domains based on their semantic similarity with respect to the input query. We further use n-gram co-occurrence distribution to recognize rare word patterns associated with specific domains. We aggregate the next word probability distribution based on the relative importance of different domains. Extensive experiments on three user-specific speech-to-text tasks for meetings, TED talks, and financial earnings calls show that DOC-RAG significantly outperforms strong baselines with an 8-15{\%} improvement in terms of perplexity and a 4-7{\%} reduction in terms of Word Error Rates in various settings.", }

We propose DOC-RAG - Domain-distributed Co-occurrence Retrieval Augmentation for ASR language model personalization aiming to improve the automatic speech recognition of rare word patterns in unseen domains. Our approach involves contrastively training a document retrieval module to rank external knowledge domains based on their semantic similarity with respect to the input query. We further use n-gram co-occurrence distribution to recognize rare word patterns associated with specific domains. We aggregate the next word probability distribution based on the relative importance of different domains. Extensive experiments on three user-specific speech-to-text tasks for meetings, TED talks, and financial earnings calls show that DOC-RAG significantly outperforms strong baselines with an 8-15{\%} improvement in terms of perplexity and a 4-7{\%} reduction in terms of Word Error Rates in various settings.

DOC-RAG: ASR Language Model Personalization with Domain-Distributed Co-occurrence Retrieval Augmentation

lrec-main.457

Poster

https://aclanthology.org/2024.lrec-main.458.bib

https://aclanthology.org/2024.lrec-main.458/

@inproceedings{mathur-etal-2024-docscript, title = "{D}oc{S}cript: Document-level Script Event Prediction", author = "Mathur, Puneet and Morariu, Vlad I. and Garimella, Aparna and Dernoncourt, Franck and Gu, Jiuxiang and Sawhney, Ramit and Nakov, Preslav and Manocha, Dinesh and Jain, Rajiv", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.458", pages = "5140--5155", abstract = "We present a novel task of document-level script event prediction, which aims to predict the next event given a candidate list of narrative events in long-form documents. To enable this, we introduce DocSEP, a challenging dataset in two new domains - contractual documents and Wikipedia articles, where timeline events may be paragraphs apart and may require multi-hop temporal and causal reasoning. We benchmark existing baselines and present a novel architecture called DocScript to learn sequential ordering between events at the document scale. Our experimental results on the DocSEP dataset demonstrate that learning longer-range dependencies between events is a key challenge and show that contemporary LLMs such as ChatGPT and FlanT5 struggle to solve this task, indicating their lack of reasoning abilities for understanding causal relationships and temporal sequences within long texts.", }

We present a novel task of document-level script event prediction, which aims to predict the next event given a candidate list of narrative events in long-form documents. To enable this, we introduce DocSEP, a challenging dataset in two new domains - contractual documents and Wikipedia articles, where timeline events may be paragraphs apart and may require multi-hop temporal and causal reasoning. We benchmark existing baselines and present a novel architecture called DocScript to learn sequential ordering between events at the document scale. Our experimental results on the DocSEP dataset demonstrate that learning longer-range dependencies between events is a key challenge and show that contemporary LLMs such as ChatGPT and FlanT5 struggle to solve this task, indicating their lack of reasoning abilities for understanding causal relationships and temporal sequences within long texts.

DocScript: Document-level Script Event Prediction

lrec-main.458

Poster

https://aclanthology.org/2024.lrec-main.459.bib

https://aclanthology.org/2024.lrec-main.459/

@inproceedings{pan-etal-2024-document, title = "Document-Level Event Extraction via Information Interaction Based on Event Relation and Argument Correlation", author = "Pan, Bangze and Li, Yang and Wang, Suge and Li, Xiaoli and Li, Deyu and Liao, Jian and Zheng, Jianxing", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.459", pages = "5156--5166", abstract = "Document-level Event Extraction (DEE) is a vital task in NLP as it seeks to automatically recognize and extract event information from a document. However, current approaches often overlook intricate relationships among events and subtle correlations among arguments within a document, which can significantly impact the effectiveness of event type recognition and the extraction of cross-sentence arguments in DEE task. This paper proposes a novel Correlation Association Interactive Network (CAINet), comprising two key components: event relationship graph and argument correlation graph. In particular, the event relationship graph models the relationship among various events through structural associations among event nodes and sentence nodes, to improve the accuracy of event recognition. On the other hand, the arguments correlation graph models the correlations among arguments by quantifying the strength of association among arguments, to effectively aggregate cross-sentence arguments, contributing to the overall success of DEE. Furthermore, we use the large language model to execute DEE task experiments. Experimental results show the proposed CAINet outperforms existing state-of-the-art models and large language models in terms of F1-score across two benchmark datasets.", }

Document-level Event Extraction (DEE) is a vital task in NLP as it seeks to automatically recognize and extract event information from a document. However, current approaches often overlook intricate relationships among events and subtle correlations among arguments within a document, which can significantly impact the effectiveness of event type recognition and the extraction of cross-sentence arguments in DEE task. This paper proposes a novel Correlation Association Interactive Network (CAINet), comprising two key components: event relationship graph and argument correlation graph. In particular, the event relationship graph models the relationship among various events through structural associations among event nodes and sentence nodes, to improve the accuracy of event recognition. On the other hand, the arguments correlation graph models the correlations among arguments by quantifying the strength of association among arguments, to effectively aggregate cross-sentence arguments, contributing to the overall success of DEE. Furthermore, we use the large language model to execute DEE task experiments. Experimental results show the proposed CAINet outperforms existing state-of-the-art models and large language models in terms of F1-score across two benchmark datasets.

Document-Level Event Extraction via Information Interaction Based on Event Relation and Argument Correlation

lrec-main.459

Poster

https://aclanthology.org/2024.lrec-main.460.bib

https://aclanthology.org/2024.lrec-main.460/

@inproceedings{zhang-etal-2024-document, title = "Document Set Expansion with Positive-Unlabeled Learning Using Intractable Density Estimation", author = "Zhang, Haiyang and Chen, Qiuyi and Zou, Yanjie and Wang, Jia and Pan, Yushan and Stevenson, Mark", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.460", pages = "5167--5173", abstract = "The Document Set Expansion (DSE) task involves identifying relevant documents from large collections based on a limited set of example documents. Previous research has highlighted Positive and Unlabeled (PU) learning as a promising approach for this task. However, most PU methods rely on the unrealistic assumption of knowing the class prior for positive samples in the collection. To address this limitation, this paper introduces a novel PU learning framework that utilizes intractable density estimation models. Experiments conducted on PubMed and Covid datasets in a transductive setting showcase the effectiveness of the proposed method for DSE. Code is available from https://github.com/Beautifuldog01/Document-set-expansion-puDE.", }

The Document Set Expansion (DSE) task involves identifying relevant documents from large collections based on a limited set of example documents. Previous research has highlighted Positive and Unlabeled (PU) learning as a promising approach for this task. However, most PU methods rely on the unrealistic assumption of knowing the class prior for positive samples in the collection. To address this limitation, this paper introduces a novel PU learning framework that utilizes intractable density estimation models. Experiments conducted on PubMed and Covid datasets in a transductive setting showcase the effectiveness of the proposed method for DSE. Code is available from https://github.com/Beautifuldog01/Document-set-expansion-puDE.

Document Set Expansion with Positive-Unlabeled Learning Using Intractable Density Estimation

lrec-main.460

Poster

https://aclanthology.org/2024.lrec-main.461.bib

https://aclanthology.org/2024.lrec-main.461/

@inproceedings{liu-etal-2024-emergent, title = "Do Emergent Abilities Exist in Quantized Large Language Models: An Empirical Study", author = "Liu, Peiyu and Liu, Zikang and Gao, Ze-Feng and Gao, Dawei and Zhao, Wayne Xin and Li, Yaliang and Ding, Bolin and Wen, Ji-Rong", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.461", pages = "5174--5190", abstract = "Despite the superior performance, Large Language Models (LLMs) require significant computational resources for deployment and use. To overcome this issue, quantization methods have been widely applied to reduce the memory footprint of LLMs as well as increase the inference rate. However, a major challenge is that low-bit quantization methods often lead to performance degradation. It is important to understand how quantization impacts the capacity of LLMs. Different from previous studies focused on overall performance, this work aims to investigate the impact of quantization on \textit{emergent abilities}, which are important characteristics that distinguish LLMs from small language models. Specifically, we examine the abilities of in-context learning, chain-of-thought reasoning, and instruction-following in quantized LLMs. Our empirical experiments show that these emergent abilities still exist in 4-bit quantization models, while 2-bit models encounter severe performance degradation on the test of these abilities. To improve the performance of low-bit models, we conduct two special experiments: (1) fine-gained impact analysis that studies which components (or substructures) are more sensitive to quantization, and (2) performance compensation through model fine-tuning. Our work derives a series of important findings to understand the impact of quantization on emergent abilities and sheds light on the possibilities of extremely low-bit quantization for LLMs.", }

Despite the superior performance, Large Language Models (LLMs) require significant computational resources for deployment and use. To overcome this issue, quantization methods have been widely applied to reduce the memory footprint of LLMs as well as increase the inference rate. However, a major challenge is that low-bit quantization methods often lead to performance degradation. It is important to understand how quantization impacts the capacity of LLMs. Different from previous studies focused on overall performance, this work aims to investigate the impact of quantization on \textit{emergent abilities}, which are important characteristics that distinguish LLMs from small language models. Specifically, we examine the abilities of in-context learning, chain-of-thought reasoning, and instruction-following in quantized LLMs. Our empirical experiments show that these emergent abilities still exist in 4-bit quantization models, while 2-bit models encounter severe performance degradation on the test of these abilities. To improve the performance of low-bit models, we conduct two special experiments: (1) fine-gained impact analysis that studies which components (or substructures) are more sensitive to quantization, and (2) performance compensation through model fine-tuning. Our work derives a series of important findings to understand the impact of quantization on emergent abilities and sheds light on the possibilities of extremely low-bit quantization for LLMs.

Do Emergent Abilities Exist in Quantized Large Language Models: An Empirical Study

lrec-main.461

Poster

2307.08072

https://aclanthology.org/2024.lrec-main.462.bib

https://aclanthology.org/2024.lrec-main.462/

@inproceedings{yuan-etal-2024-chatgpt, title = "Does {C}hat{GPT} Know That It Does Not Know? Evaluating the Black-Box Calibration of {C}hat{GPT}", author = "Yuan, Youliang and Wang, Wenxuan and Guo, Qingshuo and Xiong, Yiming and Shen, Chihao and He, Pinjia", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.462", pages = "5191--5201", abstract = "Recently, ChatGPT has demonstrated remarkable performance in various downstream tasks such as open-domain question answering, machine translation, and code generation. As a general-purpose task solver, an intriguing inquiry arises: Does ChatGPT itself know that it does not know, without any access to internal states? In response to this query, we present an initial evaluation of ChatGPT for black-box calibration. We designed three types of proxy confidence, from three perspectives to assess its performance. Experiments are conducted on five datasets, spanning four tasks, and the results show that ChatGPT has a degree of capability for black-box calibration. Specifically, proxy confidence displayed a significantly positive Pearson correlation (95.16{\%}) with accuracy in the TruthfulQA dataset, while revealing a negative correlation in the ModAr dataset. We delved deeper into ChatGPT{'}s black-box calibration ability by examining failure cases in the ModAr dataset. Our analysis revealed that ChatGPT{'}s tendency to exhibit overconfidence may stem from its reliance on semantic priors. Furthermore, we investigated why ChatGPT performs relatively well in TruthfulQA. The findings suggest that ChatGPT might implicitly acquire calibration skills during the reinforcement learning process, rather than relying solely on simplistic heuristics.", }

Recently, ChatGPT has demonstrated remarkable performance in various downstream tasks such as open-domain question answering, machine translation, and code generation. As a general-purpose task solver, an intriguing inquiry arises: Does ChatGPT itself know that it does not know, without any access to internal states? In response to this query, we present an initial evaluation of ChatGPT for black-box calibration. We designed three types of proxy confidence, from three perspectives to assess its performance. Experiments are conducted on five datasets, spanning four tasks, and the results show that ChatGPT has a degree of capability for black-box calibration. Specifically, proxy confidence displayed a significantly positive Pearson correlation (95.16{\%}) with accuracy in the TruthfulQA dataset, while revealing a negative correlation in the ModAr dataset. We delved deeper into ChatGPT{'}s black-box calibration ability by examining failure cases in the ModAr dataset. Our analysis revealed that ChatGPT{'}s tendency to exhibit overconfidence may stem from its reliance on semantic priors. Furthermore, we investigated why ChatGPT performs relatively well in TruthfulQA. The findings suggest that ChatGPT might implicitly acquire calibration skills during the reinforcement learning process, rather than relying solely on simplistic heuristics.

Does ChatGPT Know That It Does Not Know? Evaluating the Black-Box Calibration of ChatGPT

lrec-main.462

Poster

https://aclanthology.org/2024.lrec-main.463.bib

https://aclanthology.org/2024.lrec-main.463/

@inproceedings{hu-etal-2024-generator, title = "Does the Generator Mind Its Contexts? An Analysis of Generative Model Faithfulness under Context Transfer", author = "Hu, Xinshuo and Li, Dongfang and Li, Xiaoguang and Wu, Yuxiang and Shang, Lifeng and Hu, Baotian", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.463", pages = "5202--5211", abstract = "he present study introduces the knowledge-augmented generator, which is specifically designed to produce information that remains grounded in contextual knowledge, regardless of alterations in the context. Previous research has predominantly focused on examining hallucinations stemming from static input, such as in the domains of summarization or machine translation. However, our investigation delves into the faithfulness of generative question answering in the presence of dynamic knowledge. Our objective is to explore the existence of hallucinations arising from parametric memory when contextual knowledge undergoes changes, while also analyzing the underlying causes for their occurrence. In order to efficiently address this issue, we propose a straightforward yet effective measure for detecting such hallucinations. Intriguingly, our investigation uncovers that all models exhibit a tendency to generate previous answers as hallucinations. To gain deeper insights into the underlying causes of this phenomenon, we conduct a series of experiments that verify the critical role played by context in hallucination, both during training and testing, from various perspectives.", }

he present study introduces the knowledge-augmented generator, which is specifically designed to produce information that remains grounded in contextual knowledge, regardless of alterations in the context. Previous research has predominantly focused on examining hallucinations stemming from static input, such as in the domains of summarization or machine translation. However, our investigation delves into the faithfulness of generative question answering in the presence of dynamic knowledge. Our objective is to explore the existence of hallucinations arising from parametric memory when contextual knowledge undergoes changes, while also analyzing the underlying causes for their occurrence. In order to efficiently address this issue, we propose a straightforward yet effective measure for detecting such hallucinations. Intriguingly, our investigation uncovers that all models exhibit a tendency to generate previous answers as hallucinations. To gain deeper insights into the underlying causes of this phenomenon, we conduct a series of experiments that verify the critical role played by context in hallucination, both during training and testing, from various perspectives.

Does the Generator Mind Its Contexts? An Analysis of Generative Model Faithfulness under Context Transfer

lrec-main.463

Poster

2402.14488

https://aclanthology.org/2024.lrec-main.464.bib

https://aclanthology.org/2024.lrec-main.464/

@inproceedings{pucci-ranaldi-2024-language, title = "Does the Language Matter? Curriculum Learning over Neo-{L}atin Languages", author = "Pucci, Giulia and Ranaldi, Leonardo", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.464", pages = "5212--5220", abstract = "Curriculum Learning (CL) is emerging as a relevant technique to reduce the cost of pre-training Large Language Models. The idea, tested for the English language, is to train LLMs by organizing training examples from the simplest to the most complex. Complexity measures may depend on the specific language. Hence, this paper aims to investigate whether CL and the complexity measure can be easily exported to other languages. For this reason, we present a set of linguistically motivated measures to determine the complexity of examples, which has been used in English: these measures are based on text length, rarity, and comprehensibility. We then test the approach to two Romance languages: Italian and French. Our results show that the technique can be easily exported to languages other than English without adaptation.", }

Curriculum Learning (CL) is emerging as a relevant technique to reduce the cost of pre-training Large Language Models. The idea, tested for the English language, is to train LLMs by organizing training examples from the simplest to the most complex. Complexity measures may depend on the specific language. Hence, this paper aims to investigate whether CL and the complexity measure can be easily exported to other languages. For this reason, we present a set of linguistically motivated measures to determine the complexity of examples, which has been used in English: these measures are based on text length, rarity, and comprehensibility. We then test the approach to two Romance languages: Italian and French. Our results show that the technique can be easily exported to languages other than English without adaptation.

Does the Language Matter? Curriculum Learning over Neo-Latin Languages

lrec-main.464

Poster

https://aclanthology.org/2024.lrec-main.465.bib

https://aclanthology.org/2024.lrec-main.465/

@inproceedings{van-noord-etal-2024-language, title = "Do Language Models Care about Text Quality? Evaluating Web-Crawled Corpora across 11 Languages", author = "van Noord, Rik and Kuzman, Taja and Rupnik, Peter and Ljube{\v{s}}i{\'c}, Nikola and Espl{\`a}-Gomis, Miquel and Ram{\'\i}rez-S{\'a}nchez, Gema and Toral, Antonio", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.465", pages = "5221--5234", abstract = "Large, curated, web-crawled corpora play a vital role in training language models (LMs). They form the lion{'}s share of the training data in virtually all recent LMs, such as the well-known GPT, LLaMA and XLM-RoBERTa models. However, despite this importance, relatively little attention has been given to the quality of these corpora. In this paper, we compare four of the currently most relevant large, web-crawled corpora (CC100, MaCoCu, mC4 and OSCAR) across eleven lower-resourced European languages. Our approach is two-fold: first, we perform an intrinsic evaluation by performing a human evaluation of the quality of samples taken from different corpora; then, we assess the practical impact of the qualitative differences by training specific LMs on each of the corpora and evaluating their performance on downstream tasks. We find that there are clear differences in quality of the corpora, with MaCoCu and OSCAR obtaining the best results. However, during the extrinsic evaluation, we actually find that the CC100 corpus achieves the highest scores. We conclude that, in our experiments, the quality of the web-crawled corpora does not seem to play a significant role when training LMs.", }

Large, curated, web-crawled corpora play a vital role in training language models (LMs). They form the lion{'}s share of the training data in virtually all recent LMs, such as the well-known GPT, LLaMA and XLM-RoBERTa models. However, despite this importance, relatively little attention has been given to the quality of these corpora. In this paper, we compare four of the currently most relevant large, web-crawled corpora (CC100, MaCoCu, mC4 and OSCAR) across eleven lower-resourced European languages. Our approach is two-fold: first, we perform an intrinsic evaluation by performing a human evaluation of the quality of samples taken from different corpora; then, we assess the practical impact of the qualitative differences by training specific LMs on each of the corpora and evaluating their performance on downstream tasks. We find that there are clear differences in quality of the corpora, with MaCoCu and OSCAR obtaining the best results. However, during the extrinsic evaluation, we actually find that the CC100 corpus achieves the highest scores. We conclude that, in our experiments, the quality of the web-crawled corpora does not seem to play a significant role when training LMs.

Do Language Models Care about Text Quality? Evaluating Web-Crawled Corpora across 11 Languages

lrec-main.465

Poster

2403.08693

https://huggingface.co/papers/2403.08693

https://aclanthology.org/2024.lrec-main.466.bib

https://aclanthology.org/2024.lrec-main.466/

@inproceedings{perez-almendros-camacho-collados-2024-large, title = "Do Large Language Models Understand Mansplaining? Well, Actually...", author = "Perez Almendros, Carla and Camacho-Collados, Jose", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.466", pages = "5235--5246", abstract = "Gender bias has been widely studied by the NLP community. However, other more subtle variations of it, such as mansplaining, have yet received little attention. Mansplaining is a discriminatory behaviour that consists of a condescending treatment or discourse towards women. In this paper, we introduce and analyze Well, actually..., a corpus of 886 mansplaining stories experienced by women. We analyze the corpus in terms of features such as offensiveness, sentiment or misogyny, among others. We also explore to what extent Large Language Models (LLMs) can understand and identify mansplaining and other gender-related microaggressions. Specifically, we experiment with ChatGPT-3.5-Turbo and LLaMA-2 (13b and 70b), with both targeted and open questions. Our findings suggest that, although they can identify mansplaining to some extent, LLMs still struggle to point out this attitude and will even reproduce some of the social patterns behind mansplaining situations, for instance by praising men for giving unsolicited advice to women.", }

Gender bias has been widely studied by the NLP community. However, other more subtle variations of it, such as mansplaining, have yet received little attention. Mansplaining is a discriminatory behaviour that consists of a condescending treatment or discourse towards women. In this paper, we introduce and analyze Well, actually..., a corpus of 886 mansplaining stories experienced by women. We analyze the corpus in terms of features such as offensiveness, sentiment or misogyny, among others. We also explore to what extent Large Language Models (LLMs) can understand and identify mansplaining and other gender-related microaggressions. Specifically, we experiment with ChatGPT-3.5-Turbo and LLaMA-2 (13b and 70b), with both targeted and open questions. Our findings suggest that, although they can identify mansplaining to some extent, LLMs still struggle to point out this attitude and will even reproduce some of the social patterns behind mansplaining situations, for instance by praising men for giving unsolicited advice to women.

Do Large Language Models Understand Mansplaining? Well, Actually...

lrec-main.466

Poster

https://aclanthology.org/2024.lrec-main.467.bib

https://aclanthology.org/2024.lrec-main.467/

@inproceedings{li-gaussier-2024-domain, title = "Domain Adaptation for Dense Retrieval and Conversational Dense Retrieval through Self-Supervision by Meticulous Pseudo-Relevance Labeling", author = "Li, Minghan and Gaussier, Eric", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.467", pages = "5247--5259", abstract = "Recent studies have demonstrated that the ability of dense retrieval models to generalize to target domains with different distributions is limited, which contrasts with the results obtained with interaction-based models. Prior attempts to mitigate this challenge involved leveraging adversarial learning and query generation approaches, but both approaches nevertheless resulted in limited improvements. In this paper, we propose to combine the query-generation approach with a self-supervision approach in which pseudo-relevance labels are automatically generated on the target domain. To accomplish this, a T5-3B model is utilized for pseudo-positive labeling, and meticulous hard negatives are chosen. We also apply this strategy on conversational dense retrieval model for conversational search. A similar pseudo-labeling approach is used, but with the addition of a query-rewriting module to rewrite conversational queries for subsequent labeling. This proposed approach enables a model{'}s domain adaptation with real queries and documents from the target dataset. Experiments on standard dense retrieval and conversational dense retrieval models both demonstrate improvements on baseline models when they are fine-tuned on the pseudo-relevance labeled data.", }

Recent studies have demonstrated that the ability of dense retrieval models to generalize to target domains with different distributions is limited, which contrasts with the results obtained with interaction-based models. Prior attempts to mitigate this challenge involved leveraging adversarial learning and query generation approaches, but both approaches nevertheless resulted in limited improvements. In this paper, we propose to combine the query-generation approach with a self-supervision approach in which pseudo-relevance labels are automatically generated on the target domain. To accomplish this, a T5-3B model is utilized for pseudo-positive labeling, and meticulous hard negatives are chosen. We also apply this strategy on conversational dense retrieval model for conversational search. A similar pseudo-labeling approach is used, but with the addition of a query-rewriting module to rewrite conversational queries for subsequent labeling. This proposed approach enables a model{'}s domain adaptation with real queries and documents from the target dataset. Experiments on standard dense retrieval and conversational dense retrieval models both demonstrate improvements on baseline models when they are fine-tuned on the pseudo-relevance labeled data.

Domain Adaptation for Dense Retrieval and Conversational Dense Retrieval through Self-Supervision by Meticulous Pseudo-Relevance Labeling

lrec-main.467

Poster

2403.08970

https://aclanthology.org/2024.lrec-main.468.bib

https://aclanthology.org/2024.lrec-main.468/

@inproceedings{boumber-etal-2024-domain, title = "Domain-Agnostic Adapter Architecture for Deception Detection: Extensive Evaluations with the {DIF}rau{D} Benchmark", author = "Boumber, Dainis A. and Qachfar, Fatima Zahra and Verma, Rakesh", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.468", pages = "5260--5274", abstract = "Despite significant strides in training expansive transformer models, their deployment for niche tasks remains intricate. This paper delves into deception detection, assessing domain adaptation methodologies from a cross-domain lens using transformer Large Language Models (LLMs). We roll out a new corpus with roughly 100,000 honest and misleading statements in seven domains, designed to serve as a benchmark for multidomain deception detection. As a primary contribution, we present a novel parameter-efficient finetuning adapter, PreXIA, which was proposed and implemented as part of this work. The design is model-, domain- and task-agnostic, with broad applications that are not limited by the confines of deception or classification tasks. We comprehensively analyze and rigorously evaluate LLM tuning methods and our original design using the new benchmark, highlighting their strengths, pointing out weaknesses, and suggesting potential areas for improvement. The proposed adapter consistently outperforms all competition on the DIFrauD benchmark used in this study. To the best of our knowledge, it improves on the state-of-the-art in its class for the deception task. In addition, the evaluation process leads to unexpected findings that, at the very least, cast doubt on the conclusions made in some of the recently published research regarding reasoning ability{'}s unequivocal dominance over representations quality with respect to the relative contribution of each one to a model{'}s performance and predictions.", }

Despite significant strides in training expansive transformer models, their deployment for niche tasks remains intricate. This paper delves into deception detection, assessing domain adaptation methodologies from a cross-domain lens using transformer Large Language Models (LLMs). We roll out a new corpus with roughly 100,000 honest and misleading statements in seven domains, designed to serve as a benchmark for multidomain deception detection. As a primary contribution, we present a novel parameter-efficient finetuning adapter, PreXIA, which was proposed and implemented as part of this work. The design is model-, domain- and task-agnostic, with broad applications that are not limited by the confines of deception or classification tasks. We comprehensively analyze and rigorously evaluate LLM tuning methods and our original design using the new benchmark, highlighting their strengths, pointing out weaknesses, and suggesting potential areas for improvement. The proposed adapter consistently outperforms all competition on the DIFrauD benchmark used in this study. To the best of our knowledge, it improves on the state-of-the-art in its class for the deception task. In addition, the evaluation process leads to unexpected findings that, at the very least, cast doubt on the conclusions made in some of the recently published research regarding reasoning ability{'}s unequivocal dominance over representations quality with respect to the relative contribution of each one to a model{'}s performance and predictions.

Domain-Agnostic Adapter Architecture for Deception Detection: Extensive Evaluations with the DIFrauD Benchmark

lrec-main.468

Poster

https://aclanthology.org/2024.lrec-main.469.bib

https://aclanthology.org/2024.lrec-main.469/

@inproceedings{liu-etal-2024-domain, title = "Domain-aware and Co-adaptive Feature Transformation for Domain Adaption Few-shot Relation Extraction", author = "Liu, Yijun and Dai, Feifei and Gu, Xiaoyan and Zhai, Minghui and Li, Bo and Zhang, Meiou", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.469", pages = "5275--5285", abstract = "Few-shot relation extraction (FSRE) can alleviate the data scarcity problem in relation extraction. However, FSRE models often suffer a significant decline in performance when adapting to new domains. To overcome this issue, many researchers have focused on domain adaption FSRE (DAFSRE). Nevertheless, existing approaches primarily concentrate on the source domain, which makes it difficult to accurately transfer useful knowledge to the target domain. Additionally, the lack of distinction between relations further restricts the model performance. In this paper, we propose the domain-aware and co-adaptive feature transformation approach to address these issues. Specifically, we introduce a domain-aware transformation module that leverages the target domain distribution features to guide the domain-aware feature transformations. This can enhance the model{'}s adaptability across domains, leading to improved target domain performance. Furthermore, we design co-adaptive prototypical networks to perform co-adaptive feature transformation through a transformer mechanism. This results in more robust and distinguishable relation prototypes. Experiments on DAFSRE benchmark datasets demonstrate the effectiveness of our method, which outperforms existing models and achieves state-of-the-art performance.", }

Few-shot relation extraction (FSRE) can alleviate the data scarcity problem in relation extraction. However, FSRE models often suffer a significant decline in performance when adapting to new domains. To overcome this issue, many researchers have focused on domain adaption FSRE (DAFSRE). Nevertheless, existing approaches primarily concentrate on the source domain, which makes it difficult to accurately transfer useful knowledge to the target domain. Additionally, the lack of distinction between relations further restricts the model performance. In this paper, we propose the domain-aware and co-adaptive feature transformation approach to address these issues. Specifically, we introduce a domain-aware transformation module that leverages the target domain distribution features to guide the domain-aware feature transformations. This can enhance the model{'}s adaptability across domains, leading to improved target domain performance. Furthermore, we design co-adaptive prototypical networks to perform co-adaptive feature transformation through a transformer mechanism. This results in more robust and distinguishable relation prototypes. Experiments on DAFSRE benchmark datasets demonstrate the effectiveness of our method, which outperforms existing models and achieves state-of-the-art performance.

Domain-aware and Co-adaptive Feature Transformation for Domain Adaption Few-shot Relation Extraction

lrec-main.469

Poster

https://aclanthology.org/2024.lrec-main.470.bib

https://aclanthology.org/2024.lrec-main.470/

@inproceedings{wang-etal-2024-domain, title = "Domain Generalization via Causal Adjustment for Cross-Domain Sentiment Analysis", author = "Wang, Siyin and Zhou, Jie and Chen, Qin and Zhang, Qi and Gui, Tao and Huang, Xuanjing", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.470", pages = "5286--5298", abstract = "Domain adaption has been widely adapted for cross-domain sentiment analysis to transfer knowledge from the source domain to the target domain. Whereas, most methods are proposed under the assumption that the target (test) domain is known, making them fail to generalize well on unknown test data that is not always available in practice. In this paper, we focus on the problem of domain generalization for cross-domain sentiment analysis. Specifically, we propose a backdoor adjustment-based causal model to disentangle the domain-specific and domain-invariant representations that play essential roles in tackling domain shift. First, we rethink the cross-domain sentiment analysis task in a causal view to model the causal-and-effect relationships among different variables. Then, to learn an invariant feature representation, we remove the effect of domain confounders (e.g., domain knowledge) using the backdoor adjustment. A series of experiments over many homologous and diverse datasets show the great performance and robustness of our model by comparing it with the state-of-the-art domain generalization baselines.", }

Domain adaption has been widely adapted for cross-domain sentiment analysis to transfer knowledge from the source domain to the target domain. Whereas, most methods are proposed under the assumption that the target (test) domain is known, making them fail to generalize well on unknown test data that is not always available in practice. In this paper, we focus on the problem of domain generalization for cross-domain sentiment analysis. Specifically, we propose a backdoor adjustment-based causal model to disentangle the domain-specific and domain-invariant representations that play essential roles in tackling domain shift. First, we rethink the cross-domain sentiment analysis task in a causal view to model the causal-and-effect relationships among different variables. Then, to learn an invariant feature representation, we remove the effect of domain confounders (e.g., domain knowledge) using the backdoor adjustment. A series of experiments over many homologous and diverse datasets show the great performance and robustness of our model by comparing it with the state-of-the-art domain generalization baselines.

Domain Generalization via Causal Adjustment for Cross-Domain Sentiment Analysis

lrec-main.470

Poster

2402.14536

https://aclanthology.org/2024.lrec-main.471.bib

https://aclanthology.org/2024.lrec-main.471/

@inproceedings{chika-etal-2024-domain, title = "Domain Transferable Semantic Frames for Expert Interview Dialogues", author = "Chika, Taishi and Okahisa, Taro and Kodama, Takashi and Huang, Yin Jou and Murawaki, Yugo and Kurohashi, Sadao", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.471", pages = "5299--5308", abstract = "Interviews are an effective method to elicit critical skills to perform particular processes in various domains. In order to understand the knowledge structure of these domain-specific processes, we consider semantic role and predicate annotation based on Frame Semantics. We introduce a dataset of interview dialogues with experts in the culinary and gardening domains, each annotated with semantic frames. This dataset consists of (1) 308 interview dialogues related to the culinary domain, originally assembled by Okahisa et al. (2022), and (2) 100 interview dialogues associated with the gardening domain, which we newly acquired. The labeling specifications take into account the domain-transferability by adopting domain-agnostic labels for frame elements. In addition, we conducted domain transfer experiments from the culinary domain to the gardening domain to examine the domain transferability with our dataset. The experimental results showed the effectiveness of our domain-agnostic labeling scheme.", }

Interviews are an effective method to elicit critical skills to perform particular processes in various domains. In order to understand the knowledge structure of these domain-specific processes, we consider semantic role and predicate annotation based on Frame Semantics. We introduce a dataset of interview dialogues with experts in the culinary and gardening domains, each annotated with semantic frames. This dataset consists of (1) 308 interview dialogues related to the culinary domain, originally assembled by Okahisa et al. (2022), and (2) 100 interview dialogues associated with the gardening domain, which we newly acquired. The labeling specifications take into account the domain-transferability by adopting domain-agnostic labels for frame elements. In addition, we conducted domain transfer experiments from the culinary domain to the gardening domain to examine the domain transferability with our dataset. The experimental results showed the effectiveness of our domain-agnostic labeling scheme.

Domain Transferable Semantic Frames for Expert Interview Dialogues

lrec-main.471

Poster

https://aclanthology.org/2024.lrec-main.472.bib

https://aclanthology.org/2024.lrec-main.472/

@inproceedings{rodriguez-etal-2024-neural, title = "Do Neural Language Models Inferentially Compose Concepts the Way Humans Can?", author = {Rodriguez, Amilleah and Wang, Shaonan and Pylkk{\"a}nen, Liina}, editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.472", pages = "5309--5314", abstract = "While compositional interpretation is the core of language understanding, humans also derive meaning via inference. For example, while the phrase {``}the blue hat{''} introduces a blue hat into the discourse via the direct composition of {``}blue{''} and {``}hat,{''} the same discourse entity is introduced by the phrase {``}the blue color of this hat{''} despite the absence of any local composition between {``}blue{''} and {``}hat.{''} Instead, we infer that if the color is blue and it belongs to the hat, the hat must be blue. We tested the performance of neural language models and humans on such inferentially driven conceptual compositions, eliciting probability estimates for a noun in a minimally composed phrase, {``}This blue hat{''}, following contexts that had introduced the conceptual combinations of those nouns and adjectives either syntactically or inferentially. Surprisingly, our findings reveal significant disparities between the performance of neural language models and human judgments. Among the eight models evaluated, RoBERTa, BERT-large, and GPT-2 exhibited the closest resemblance to human responses, while other models faced challenges in accurately identifying compositions in the provided contexts. Our study reveals that language models and humans may rely on different approaches to represent and compose lexical items across sentence structure. All data and code are accessible at https://github.com/wangshaonan/BlueHat.", }

While compositional interpretation is the core of language understanding, humans also derive meaning via inference. For example, while the phrase {``}the blue hat{''} introduces a blue hat into the discourse via the direct composition of {``}blue{''} and {``}hat,{''} the same discourse entity is introduced by the phrase {``}the blue color of this hat{''} despite the absence of any local composition between {``}blue{''} and {``}hat.{''} Instead, we infer that if the color is blue and it belongs to the hat, the hat must be blue. We tested the performance of neural language models and humans on such inferentially driven conceptual compositions, eliciting probability estimates for a noun in a minimally composed phrase, {``}This blue hat{''}, following contexts that had introduced the conceptual combinations of those nouns and adjectives either syntactically or inferentially. Surprisingly, our findings reveal significant disparities between the performance of neural language models and human judgments. Among the eight models evaluated, RoBERTa, BERT-large, and GPT-2 exhibited the closest resemblance to human responses, while other models faced challenges in accurately identifying compositions in the provided contexts. Our study reveals that language models and humans may rely on different approaches to represent and compose lexical items across sentence structure. All data and code are accessible at https://github.com/wangshaonan/BlueHat.

Do Neural Language Models Inferentially Compose Concepts the Way Humans Can?

lrec-main.472

Poster

https://aclanthology.org/2024.lrec-main.473.bib

https://aclanthology.org/2024.lrec-main.473/

@inproceedings{dimas-furtado-etal-2024-dore, title = "{DORE}: A Dataset for {P}ortuguese Definition Generation", author = "Dimas Furtado, Anna Beatriz and Ranasinghe, Tharindu and Blain, Frederic and Mitkov, Ruslan", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.473", pages = "5315--5322", abstract = "Definition modelling (DM) is the task of automatically generating a dictionary definition of a specific word. Computational systems that are capable of DM can have numerous applications benefiting a wide range of audiences. As DM is considered a supervised natural language generation problem, these systems require large annotated datasets to train the machine learning (ML) models. Several DM datasets have been released for English and other high-resource languages. While Portuguese is considered a mid/high-resource language in most natural language processing tasks and is spoken by more than 200 million native speakers, there is no DM dataset available for Portuguese. In this research, we fill this gap by introducing DORE; the first dataset for \textbf{D}efinition M\textbf{O}delling for Po\textbf{R}tugu\textbf{E}se containing more than 100,000 definitions. We also evaluate several deep learning based DM models on DORE and report the results. The dataset and the findings of this paper will facilitate research and study of Portuguese in wider contexts.", }

Definition modelling (DM) is the task of automatically generating a dictionary definition of a specific word. Computational systems that are capable of DM can have numerous applications benefiting a wide range of audiences. As DM is considered a supervised natural language generation problem, these systems require large annotated datasets to train the machine learning (ML) models. Several DM datasets have been released for English and other high-resource languages. While Portuguese is considered a mid/high-resource language in most natural language processing tasks and is spoken by more than 200 million native speakers, there is no DM dataset available for Portuguese. In this research, we fill this gap by introducing DORE; the first dataset for \textbf{D}efinition M\textbf{O}delling for Po\textbf{R}tugu\textbf{E}se containing more than 100,000 definitions. We also evaluate several deep learning based DM models on DORE and report the results. The dataset and the findings of this paper will facilitate research and study of Portuguese in wider contexts.

DORE: A Dataset for Portuguese Definition Generation

lrec-main.473

Poster

2403.18018

https://aclanthology.org/2024.lrec-main.474.bib

https://aclanthology.org/2024.lrec-main.474/

@inproceedings{seth-etal-2024-dosa, title = "{DOSA}: A Dataset of Social Artifacts from Different {I}ndian Geographical Subcultures", author = "Seth, Agrima and Ahuja, Sanchit and Bali, Kalika and Sitaram, Sunayana", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.474", pages = "5323--5337", abstract = "Generative models are increasingly being used in various applications, such as text generation, commonsense reasoning, and question-answering. To be effective globally, these models must be aware of and account for local socio-cultural contexts, making it necessary to have benchmarks to evaluate the models for their cultural familiarity. Since the training data for LLMs is web-based and the Web is limited in its representation of information, it does not capture knowledge present within communities that are not on the Web. Thus, these models exacerbate the inequities, semantic misalignment, and stereotypes from the Web. There has been a growing call for community-centered participatory research methods in NLP. In this work, we respond to this call by using participatory research methods to introduce DOSA, the first community-generated Dataset of 615 Social Artifacts, by engaging with 260 participants from 19 different Indian geographic subcultures. We use a gamified framework that relies on collective sensemaking to collect the names and descriptions of these artifacts such that the descriptions semantically align with the shared sensibilities of the individuals from those cultures. Next, we benchmark four popular LLMs and find that they show significant variation across regional sub-cultures in their ability to infer the artifacts.", }

Generative models are increasingly being used in various applications, such as text generation, commonsense reasoning, and question-answering. To be effective globally, these models must be aware of and account for local socio-cultural contexts, making it necessary to have benchmarks to evaluate the models for their cultural familiarity. Since the training data for LLMs is web-based and the Web is limited in its representation of information, it does not capture knowledge present within communities that are not on the Web. Thus, these models exacerbate the inequities, semantic misalignment, and stereotypes from the Web. There has been a growing call for community-centered participatory research methods in NLP. In this work, we respond to this call by using participatory research methods to introduce DOSA, the first community-generated Dataset of 615 Social Artifacts, by engaging with 260 participants from 19 different Indian geographic subcultures. We use a gamified framework that relies on collective sensemaking to collect the names and descriptions of these artifacts such that the descriptions semantically align with the shared sensibilities of the individuals from those cultures. Next, we benchmark four popular LLMs and find that they show significant variation across regional sub-cultures in their ability to infer the artifacts.

DOSA: A Dataset of Social Artifacts from Different Indian Geographical Subcultures

lrec-main.474

Poster

2403.14651

https://aclanthology.org/2024.lrec-main.475.bib

https://aclanthology.org/2024.lrec-main.475/

@inproceedings{huang-etal-2024-dp, title = "{DP}-{CRE}: Continual Relation Extraction via Decoupled Contrastive Learning and Memory Structure Preservation", author = "Huang, Mengyi and Xiao, Meng and Wang, Ludi and Du, Yi", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.475", pages = "5338--5349", abstract = "Continuous Relation Extraction (CRE) aims to incrementally learn relation knowledge from a non-stationary stream of data. Since the introduction of new relational tasks can overshadow previously learned information, catastrophic forgetting becomes a significant challenge in this domain. Current replay-based training paradigms prioritize all data uniformly and train memory samples through multiple rounds, which would result in overfitting old tasks and pronounced bias towards new tasks because of the imbalances of the replay set. To handle the problem, we introduce the DecouPled CRE (DP-CRE) framework that decouples the process of prior information preservation and new knowledge acquisition. This framework examines alterations in the embedding space as new relation classes emerge, distinctly managing the preservation and acquisition of knowledge. Extensive experiments show that DP-CRE significantly outperforms other CRE baselines across two datasets.", }

Continuous Relation Extraction (CRE) aims to incrementally learn relation knowledge from a non-stationary stream of data. Since the introduction of new relational tasks can overshadow previously learned information, catastrophic forgetting becomes a significant challenge in this domain. Current replay-based training paradigms prioritize all data uniformly and train memory samples through multiple rounds, which would result in overfitting old tasks and pronounced bias towards new tasks because of the imbalances of the replay set. To handle the problem, we introduce the DecouPled CRE (DP-CRE) framework that decouples the process of prior information preservation and new knowledge acquisition. This framework examines alterations in the embedding space as new relation classes emerge, distinctly managing the preservation and acquisition of knowledge. Extensive experiments show that DP-CRE significantly outperforms other CRE baselines across two datasets.

DP-CRE: Continual Relation Extraction via Decoupled Contrastive Learning and Memory Structure Preservation

lrec-main.475

Poster

2403.02718

https://aclanthology.org/2024.lrec-main.476.bib

https://aclanthology.org/2024.lrec-main.476/

@inproceedings{gao-etal-2024-dr3, title = "Dr3: Ask Large Language Models Not to Give Off-Topic Answers in Open Domain Multi-Hop Question Answering", author = "Gao, Yuan and Zhu, Yiheng and Cao, Yuanbin and Zhou, Yinzhi and Wu, Zhen and Chen, Yujie and Wu, Shenglan and Hu, Haoyuan and Dai, Xinyu", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.476", pages = "5350--5364", abstract = "Open Domain Multi-Hop Question Answering (ODMHQA) plays a crucial role in Natural Language Processing (NLP) by aiming to answer complex questions through multi-step reasoning over retrieved information from external knowledge sources. Recently, Large Language Models (LLMs) have demonstrated remarkable performance in solving ODMHQA owing to their capabilities including planning, reasoning, and utilizing tools. However, LLMs may generate off-topic answers when attempting to solve ODMHQA, namely the generated answers are irrelevant to the original questions. This issue of off-topic answers accounts for approximately one-third of incorrect answers, yet remains underexplored despite its significance. To alleviate this issue, we propose the Discriminate→Re-Compose→Re- Solve→Re-Decompose (Dr3) mechanism. Specifically, the Discriminator leverages the intrinsic capabilities of LLMs to judge whether the generated answers are off-topic. In cases where an off-topic answer is detected, the Corrector performs step-wise revisions along the reversed reasoning chain (Re-Compose→Re-Solve→Re-Decompose) until the final answer becomes on-topic. Experimental results on the HotpotQA and 2WikiMultiHopQA datasets demonstrate that our Dr3 mechanism considerably reduces the occurrence of off-topic answers in ODMHQA by nearly 13{\%}, improving the performance in Exact Match (EM) by nearly 3{\%} compared to the baseline method without the Dr3 mechanism.", }

Open Domain Multi-Hop Question Answering (ODMHQA) plays a crucial role in Natural Language Processing (NLP) by aiming to answer complex questions through multi-step reasoning over retrieved information from external knowledge sources. Recently, Large Language Models (LLMs) have demonstrated remarkable performance in solving ODMHQA owing to their capabilities including planning, reasoning, and utilizing tools. However, LLMs may generate off-topic answers when attempting to solve ODMHQA, namely the generated answers are irrelevant to the original questions. This issue of off-topic answers accounts for approximately one-third of incorrect answers, yet remains underexplored despite its significance. To alleviate this issue, we propose the Discriminate→Re-Compose→Re- Solve→Re-Decompose (Dr3) mechanism. Specifically, the Discriminator leverages the intrinsic capabilities of LLMs to judge whether the generated answers are off-topic. In cases where an off-topic answer is detected, the Corrector performs step-wise revisions along the reversed reasoning chain (Re-Compose→Re-Solve→Re-Decompose) until the final answer becomes on-topic. Experimental results on the HotpotQA and 2WikiMultiHopQA datasets demonstrate that our Dr3 mechanism considerably reduces the occurrence of off-topic answers in ODMHQA by nearly 13{\%}, improving the performance in Exact Match (EM) by nearly 3{\%} compared to the baseline method without the Dr3 mechanism.

Dr3: Ask Large Language Models Not to Give Off-Topic Answers in Open Domain Multi-Hop Question Answering

lrec-main.476

Poster

2403.12393

https://aclanthology.org/2024.lrec-main.477.bib

https://aclanthology.org/2024.lrec-main.477/

@inproceedings{yuan-etal-2024-drama, title = "{DRAMA}: Dynamic Multi-Granularity Graph Estimate Retrieval over Tabular and Textual Question Answering", author = "Yuan, Ruize and Ao, Xiang and Zeng, Li and He, Qing", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.477", pages = "5365--5375", abstract = "The TableTextQA task requires finding the answer to the question from a combination of tabular and textual data, which has been gaining increasing attention. The row-based approaches have demonstrated remarkable effectiveness. However, they suffer from the following limitations: (1) a lack of interaction between rows; (2) excessively long input lengths; and (3) question attention shifts in the multi-hop QA task. To this end, we propose a novel method: Dynamic Multi-Granularity Graph Estimate Retrieval - DRAMA. Our method incorporates an interaction mechanism among multiple rows. Specifically, we utilize a memory bank to store the features of each row, thereby facilitating the construction of a heterogeneous graph with multi-row information. Besides, a Dynamic Graph Attention Network (DGAT) module is engaged to gauge the attention shift in the multi-hop question and eliminate the noise information dynamically. Empirical results on the widely used HybridQA and TabFact datasets demonstrate that the proposed model is effective.", }

The TableTextQA task requires finding the answer to the question from a combination of tabular and textual data, which has been gaining increasing attention. The row-based approaches have demonstrated remarkable effectiveness. However, they suffer from the following limitations: (1) a lack of interaction between rows; (2) excessively long input lengths; and (3) question attention shifts in the multi-hop QA task. To this end, we propose a novel method: Dynamic Multi-Granularity Graph Estimate Retrieval - DRAMA. Our method incorporates an interaction mechanism among multiple rows. Specifically, we utilize a memory bank to store the features of each row, thereby facilitating the construction of a heterogeneous graph with multi-row information. Besides, a Dynamic Graph Attention Network (DGAT) module is engaged to gauge the attention shift in the multi-hop question and eliminate the noise information dynamically. Empirical results on the widely used HybridQA and TabFact datasets demonstrate that the proposed model is effective.

DRAMA: Dynamic Multi-Granularity Graph Estimate Retrieval over Tabular and Textual Question Answering

lrec-main.477

Poster

https://aclanthology.org/2024.lrec-main.478.bib

https://aclanthology.org/2024.lrec-main.478/

@inproceedings{labrak-etal-2024-drbenchmark, title = "{D}r{B}enchmark: A Large Language Understanding Evaluation Benchmark for {F}rench Biomedical Domain", author = "Labrak, Yanis and Bazoge, Adrien and El Khettari, Oumaima and Rouvier, Mickael and Constant Dit Beaufils, Pacome and Grabar, Natalia and Daille, B{\'e}atrice and Quiniou, Solen and Morin, Emmanuel and Gourraud, Pierre-Antoine and Dufour, Richard", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.478", pages = "5376--5390", abstract = "The biomedical domain has sparked a significant interest in the field of Natural Language Processing (NLP), which has seen substantial advancements with pre-trained language models (PLMs). However, comparing these models has proven challenging due to variations in evaluation protocols across different models. A fair solution is to aggregate diverse downstream tasks into a benchmark, allowing for the assessment of intrinsic PLMs qualities from various perspectives. Although still limited to few languages, this initiative has been undertaken in the biomedical field, notably English and Chinese. This limitation hampers the evaluation of the latest French biomedical models, as they are either assessed on a minimal number of tasks with non-standardized protocols or evaluated using general downstream tasks. To bridge this research gap and account for the unique sensitivities of French, we present the first-ever publicly available French biomedical language understanding benchmark called DrBenchmark. It encompasses 20 diversified tasks, including named-entity recognition, part-of-speech tagging, question-answering, semantic textual similarity, or classification. We evaluate 8 state-of-the-art pre-trained masked language models (MLMs) on general and biomedical-specific data, as well as English specific MLMs to assess their cross-lingual capabilities. Our experiments reveal that no single model excels across all tasks, while generalist models are sometimes still competitive.", }

The biomedical domain has sparked a significant interest in the field of Natural Language Processing (NLP), which has seen substantial advancements with pre-trained language models (PLMs). However, comparing these models has proven challenging due to variations in evaluation protocols across different models. A fair solution is to aggregate diverse downstream tasks into a benchmark, allowing for the assessment of intrinsic PLMs qualities from various perspectives. Although still limited to few languages, this initiative has been undertaken in the biomedical field, notably English and Chinese. This limitation hampers the evaluation of the latest French biomedical models, as they are either assessed on a minimal number of tasks with non-standardized protocols or evaluated using general downstream tasks. To bridge this research gap and account for the unique sensitivities of French, we present the first-ever publicly available French biomedical language understanding benchmark called DrBenchmark. It encompasses 20 diversified tasks, including named-entity recognition, part-of-speech tagging, question-answering, semantic textual similarity, or classification. We evaluate 8 state-of-the-art pre-trained masked language models (MLMs) on general and biomedical-specific data, as well as English specific MLMs to assess their cross-lingual capabilities. Our experiments reveal that no single model excels across all tasks, while generalist models are sometimes still competitive.

DrBenchmark: A Large Language Understanding Evaluation Benchmark for French Biomedical Domain

lrec-main.478

Poster

2402.13432

https://aclanthology.org/2024.lrec-main.479.bib

https://aclanthology.org/2024.lrec-main.479/

@inproceedings{dong-etal-2024-dual, title = "Dual Complex Number Knowledge Graph Embeddings", author = "Dong, Yao and Kong, Qingchao and Wang, Lei and Luo, Yin", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.479", pages = "5391--5400", abstract = "Knowledge graph embedding, which aims to learn representations of entities and relations in large scale knowledge graphs, plays a crucial part in various downstream applications. The performance of knowledge graph embedding models mainly depends on the ability of modeling relation patterns, such as symmetry/antisymmetry, inversion and composition (commutative composition and non-commutative composition). Most existing methods fail in modeling the non-commutative composition patterns. Several methods support this kind of pattern by modeling in quaternion space or dihedral group. However, extending to such sophisticated spaces leads to a substantial increase in the amount of parameters, which greatly reduces the parameter efficiency. In this paper, we propose a new knowledge graph embedding method called dual complex number knowledge graph embeddings (DCNE), which maps entities to the dual complex number space, and represents relations as rotations in 2D space via dual complex number multiplication. The non-commutativity of the dual complex number multiplication empowers DCNE to model the non-commutative composition patterns. In the meantime, modeling relations as rotations in 2D space can effectively improve the parameter efficiency. Extensive experiments on multiple benchmark knowledge graphs empirically show that DCNE achieves significant performance in link prediction and path query answering.", }

Knowledge graph embedding, which aims to learn representations of entities and relations in large scale knowledge graphs, plays a crucial part in various downstream applications. The performance of knowledge graph embedding models mainly depends on the ability of modeling relation patterns, such as symmetry/antisymmetry, inversion and composition (commutative composition and non-commutative composition). Most existing methods fail in modeling the non-commutative composition patterns. Several methods support this kind of pattern by modeling in quaternion space or dihedral group. However, extending to such sophisticated spaces leads to a substantial increase in the amount of parameters, which greatly reduces the parameter efficiency. In this paper, we propose a new knowledge graph embedding method called dual complex number knowledge graph embeddings (DCNE), which maps entities to the dual complex number space, and represents relations as rotations in 2D space via dual complex number multiplication. The non-commutativity of the dual complex number multiplication empowers DCNE to model the non-commutative composition patterns. In the meantime, modeling relations as rotations in 2D space can effectively improve the parameter efficiency. Extensive experiments on multiple benchmark knowledge graphs empirically show that DCNE achieves significant performance in link prediction and path query answering.

Dual Complex Number Knowledge Graph Embeddings

lrec-main.479

Poster

https://aclanthology.org/2024.lrec-main.480.bib

https://aclanthology.org/2024.lrec-main.480/

@inproceedings{zhao-etal-2024-dual, title = "Dual Encoder: Exploiting the Potential of Syntactic and Semantic for Aspect Sentiment Triplet Extraction", author = "Zhao, Xiaowei and Zhou, Yong and Xu, Xiujuan", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.480", pages = "5401--5413", abstract = "Aspect Sentiment Triple Extraction (ASTE) is an emerging task in fine-grained sentiment analysis. Recent studies have employed Graph Neural Networks (GNN) to model the syntax-semantic relationships inherent in triplet elements. However, they have yet to fully tap into the vast potential of syntactic and semantic information within the ASTE task. In this work, we propose a \textit{Dual Encoder: Exploiting the potential of Syntactic and Semantic} model (D2E2S), which maximizes the syntactic and semantic relationships among words. Specifically, our model utilizes a dual-channel encoder with a BERT channel to capture semantic information, and an enhanced LSTM channel for comprehensive syntactic information capture. Subsequently, we introduce the heterogeneous feature interaction module to capture intricate interactions between dependency syntax and attention semantics, and to dynamically select vital nodes. We leverage the synergy of these modules to harness the significant potential of syntactic and semantic information in ASTE tasks. Testing on public benchmarks, our D2E2S model surpasses the current state-of-the-art(SOTA), demonstrating its effectiveness.", }

Aspect Sentiment Triple Extraction (ASTE) is an emerging task in fine-grained sentiment analysis. Recent studies have employed Graph Neural Networks (GNN) to model the syntax-semantic relationships inherent in triplet elements. However, they have yet to fully tap into the vast potential of syntactic and semantic information within the ASTE task. In this work, we propose a \textit{Dual Encoder: Exploiting the potential of Syntactic and Semantic} model (D2E2S), which maximizes the syntactic and semantic relationships among words. Specifically, our model utilizes a dual-channel encoder with a BERT channel to capture semantic information, and an enhanced LSTM channel for comprehensive syntactic information capture. Subsequently, we introduce the heterogeneous feature interaction module to capture intricate interactions between dependency syntax and attention semantics, and to dynamically select vital nodes. We leverage the synergy of these modules to harness the significant potential of syntactic and semantic information in ASTE tasks. Testing on public benchmarks, our D2E2S model surpasses the current state-of-the-art(SOTA), demonstrating its effectiveness.

Dual Encoder: Exploiting the Potential of Syntactic and Semantic for Aspect Sentiment Triplet Extraction

lrec-main.480

Poster

2402.15370

https://aclanthology.org/2024.lrec-main.481.bib

https://aclanthology.org/2024.lrec-main.481/

@inproceedings{luo-etal-2024-duetsim, title = "{D}uet{S}im: Building User Simulator with Dual Large Language Models for Task-Oriented Dialogues", author = "Luo, Xiang and Tang, Zhiwen and Wang, Jin and Zhang, Xuejie", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.481", pages = "5414--5424", abstract = "User Simulators play a pivotal role in training and evaluating task-oriented dialogue systems. Traditional user simulators typically rely on human-engineered agendas, resulting in generated responses that often lack diversity and spontaneity. Although large language models (LLMs) exhibit a remarkable capacity for generating coherent and contextually appropriate utterances, they may fall short when tasked with generating responses that effectively guide users towards their goals, particularly in dialogues with intricate constraints and requirements. This paper introduces DuetSim, a novel framework designed to address the intricate demands of task-oriented dialogues by leveraging LLMs. DuetSim stands apart from conventional approaches by employing two LLMs in tandem: one dedicated to response generation and the other focused on verification. This dual LLM approach empowers DuetSim to produce responses that not only exhibit diversity but also demonstrate accuracy and are preferred by human users. We validate the efficacy of our method through extensive experiments conducted on the MultiWOZ dataset, highlighting improvements in response quality and correctness, largely attributed to the incorporation of the second LLM.", }

User Simulators play a pivotal role in training and evaluating task-oriented dialogue systems. Traditional user simulators typically rely on human-engineered agendas, resulting in generated responses that often lack diversity and spontaneity. Although large language models (LLMs) exhibit a remarkable capacity for generating coherent and contextually appropriate utterances, they may fall short when tasked with generating responses that effectively guide users towards their goals, particularly in dialogues with intricate constraints and requirements. This paper introduces DuetSim, a novel framework designed to address the intricate demands of task-oriented dialogues by leveraging LLMs. DuetSim stands apart from conventional approaches by employing two LLMs in tandem: one dedicated to response generation and the other focused on verification. This dual LLM approach empowers DuetSim to produce responses that not only exhibit diversity but also demonstrate accuracy and are preferred by human users. We validate the efficacy of our method through extensive experiments conducted on the MultiWOZ dataset, highlighting improvements in response quality and correctness, largely attributed to the incorporation of the second LLM.

DuetSim: Building User Simulator with Dual Large Language Models for Task-Oriented Dialogues

lrec-main.481

Poster

2405.13028

https://aclanthology.org/2024.lrec-main.482.bib

https://aclanthology.org/2024.lrec-main.482/

@inproceedings{bu-etal-2024-dynamic, title = "Dynamic Knowledge Prompt for Chest {X}-ray Report Generation", author = "Bu, Shenshen and Song, Yujie and Li, Taiji and Dai, Zhiming", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.482", pages = "5425--5436", abstract = "Automatic generation of radiology reports can relieve the burden of radiologist. In the radiology library, the biased dataset and the sparse features of chest X-ray image make it difficult to generate reports. Many approaches strive to integrate prior information to enhance generation, but they fail to dynamically utilize pulmonary lesion knowledge at the instance-level. To alleviate above problem, we propose a novel Dynamic Knowledge Prompt (DKP) framework for chest X-ray report generation. The DKP can dynamically incorporate the pulmonary lesion information at the instance-level to facilitate report generation. Initially, we design a knowledge prompt for each pulmonary lesion using numerous radiology reports. After that, the DKP using an anomaly detector generates the dynamic knowledge prompt by extracting discriminative lesion features in the corresponding X-ray image. Finally, the knowledge prompt is encoded and fused with hidden states extracted from decoder, to form multi-modal features that guide visual features to generate reports. Extensive experiments on the public datasets MIMIC-CXR and IU X-Ray show that our approach achieves state-of-the-art performance.", }

Automatic generation of radiology reports can relieve the burden of radiologist. In the radiology library, the biased dataset and the sparse features of chest X-ray image make it difficult to generate reports. Many approaches strive to integrate prior information to enhance generation, but they fail to dynamically utilize pulmonary lesion knowledge at the instance-level. To alleviate above problem, we propose a novel Dynamic Knowledge Prompt (DKP) framework for chest X-ray report generation. The DKP can dynamically incorporate the pulmonary lesion information at the instance-level to facilitate report generation. Initially, we design a knowledge prompt for each pulmonary lesion using numerous radiology reports. After that, the DKP using an anomaly detector generates the dynamic knowledge prompt by extracting discriminative lesion features in the corresponding X-ray image. Finally, the knowledge prompt is encoded and fused with hidden states extracted from decoder, to form multi-modal features that guide visual features to generate reports. Extensive experiments on the public datasets MIMIC-CXR and IU X-Ray show that our approach achieves state-of-the-art performance.

Dynamic Knowledge Prompt for Chest X-ray Report Generation

lrec-main.482

Poster

https://aclanthology.org/2024.lrec-main.483.bib

https://aclanthology.org/2024.lrec-main.483/

@inproceedings{min-etal-2024-dynamic, title = "Dynamic Reward Adjustment in Multi-Reward Reinforcement Learning for Counselor Reflection Generation", author = "Min, Do June and Perez-Rosas, Veronica and Resnicow, Ken and Mihalcea, Rada", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.483", pages = "5437--5449", abstract = "In this paper, we study the problem of multi-reward reinforcement learning to jointly optimize for multiple text qualities for natural language generation. We focus on the task of counselor reflection generation, where we optimize the generators to simultaneously improve the fluency, coherence, and reflection quality of generated counselor responses. We introduce two novel bandit methods, DynaOpt and C-DynaOpt, which rely on the broad strategy of combining rewards into a single value and optimizing them simultaneously. Specifically, we employ non-contextual and contextual multi-arm bandits to dynamically adjust multiple reward weights during training. Through automatic and manual evaluations, we show that our proposed techniques, DynaOpt and C-DynaOpt, outperform existing naive and bandit baselines, showcasing their potential for enhancing language models.", }

In this paper, we study the problem of multi-reward reinforcement learning to jointly optimize for multiple text qualities for natural language generation. We focus on the task of counselor reflection generation, where we optimize the generators to simultaneously improve the fluency, coherence, and reflection quality of generated counselor responses. We introduce two novel bandit methods, DynaOpt and C-DynaOpt, which rely on the broad strategy of combining rewards into a single value and optimizing them simultaneously. Specifically, we employ non-contextual and contextual multi-arm bandits to dynamically adjust multiple reward weights during training. Through automatic and manual evaluations, we show that our proposed techniques, DynaOpt and C-DynaOpt, outperform existing naive and bandit baselines, showcasing their potential for enhancing language models.

Dynamic Reward Adjustment in Multi-Reward Reinforcement Learning for Counselor Reflection Generation

lrec-main.483

Poster

2403.13578

https://aclanthology.org/2024.lrec-main.484.bib

https://aclanthology.org/2024.lrec-main.484/

@inproceedings{hu-etal-2024-dynamic, title = "Dynamic Spatial-Temporal Aggregation for Skeleton-Aware Sign Language Recognition", author = "Hu, Lianyu and Gao, Liqing and Liu, Zekang and Feng, Wei", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.484", pages = "5450--5460", abstract = "Skeleton-aware sign language recognition (SLR) has gained popularity due to its ability to remain unaffected by background information and its lower computational requirements. Current methods utilize spatial graph modules and temporal modules to capture spatial and temporal features, respectively. However, their spatial graph modules are typically built on fixed graph structures such as graph convolutional networks or a single learnable graph, which only partially explore joint relationships. Additionally, a simple temporal convolution kernel is used to capture temporal information, which may not fully capture the complex movement patterns of different signers. To overcome these limitations, we propose a new spatial architecture consisting of two concurrent branches, which build input-sensitive joint relationships and incorporates specific domain knowledge for recognition, respectively. These two branches are followed by an aggregation process to distinguishe important joint connections. We then propose a new temporal module to model multi-scale temporal information to capture complex human dynamics. Our method achieves state-of-the-art accuracy compared to previous skeleton-aware methods on four large-scale SLR benchmarks. Moreover, our method demonstrates superior accuracy compared to RGB-based methods in most cases while requiring much fewer computational resources, bringing better accuracy-computation trade-off. Code is available at https://github.com/hulianyuyy/DSTA-SLR.", }

Skeleton-aware sign language recognition (SLR) has gained popularity due to its ability to remain unaffected by background information and its lower computational requirements. Current methods utilize spatial graph modules and temporal modules to capture spatial and temporal features, respectively. However, their spatial graph modules are typically built on fixed graph structures such as graph convolutional networks or a single learnable graph, which only partially explore joint relationships. Additionally, a simple temporal convolution kernel is used to capture temporal information, which may not fully capture the complex movement patterns of different signers. To overcome these limitations, we propose a new spatial architecture consisting of two concurrent branches, which build input-sensitive joint relationships and incorporates specific domain knowledge for recognition, respectively. These two branches are followed by an aggregation process to distinguishe important joint connections. We then propose a new temporal module to model multi-scale temporal information to capture complex human dynamics. Our method achieves state-of-the-art accuracy compared to previous skeleton-aware methods on four large-scale SLR benchmarks. Moreover, our method demonstrates superior accuracy compared to RGB-based methods in most cases while requiring much fewer computational resources, bringing better accuracy-computation trade-off. Code is available at https://github.com/hulianyuyy/DSTA-SLR.

Dynamic Spatial-Temporal Aggregation for Skeleton-Aware Sign Language Recognition

lrec-main.484

Poster

2403.12519

https://aclanthology.org/2024.lrec-main.485.bib

https://aclanthology.org/2024.lrec-main.485/

@inproceedings{grasso-etal-2024-ecoverse, title = "{E}co{V}erse: An Annotated {T}witter Dataset for Eco-Relevance Classification, Environmental Impact Analysis, and Stance Detection", author = "Grasso, Francesca and Locci, Stefano and Siragusa, Giovanni and Di Caro, Luigi", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.485", pages = "5461--5472", abstract = "Anthropogenic ecological crisis constitutes a significant challenge that all within the academy must urgently face, including the Natural Language Processing (NLP) community. While recent years have seen increasing work revolving around climate-centric discourse, crucial environmental and ecological topics outside of climate change remain largely unaddressed, despite their prominent importance. Mainstream NLP tasks, such as sentiment analysis, dominate the scene, but there remains an untouched space in the literature involving the analysis of environmental impacts of certain events and practices. To address this gap, this paper presents EcoVerse, an annotated English Twitter dataset of 3,023 tweets spanning a wide spectrum of environmental topics. We propose a three-level annotation scheme designed for Eco-Relevance Classification, Stance Detection, and introducing an original approach for Environmental Impact Analysis. We detail the data collection, filtering, and labeling process that led to the creation of the dataset. Remarkable Inter-Annotator Agreement indicates that the annotation scheme produces consistent annotations of high quality. Subsequent classification experiments using BERT-based models, including ClimateBERT, are presented. These yield encouraging results, while also indicating room for a model specifically tailored for environmental texts. The dataset is made freely available to stimulate further research.", }

Anthropogenic ecological crisis constitutes a significant challenge that all within the academy must urgently face, including the Natural Language Processing (NLP) community. While recent years have seen increasing work revolving around climate-centric discourse, crucial environmental and ecological topics outside of climate change remain largely unaddressed, despite their prominent importance. Mainstream NLP tasks, such as sentiment analysis, dominate the scene, but there remains an untouched space in the literature involving the analysis of environmental impacts of certain events and practices. To address this gap, this paper presents EcoVerse, an annotated English Twitter dataset of 3,023 tweets spanning a wide spectrum of environmental topics. We propose a three-level annotation scheme designed for Eco-Relevance Classification, Stance Detection, and introducing an original approach for Environmental Impact Analysis. We detail the data collection, filtering, and labeling process that led to the creation of the dataset. Remarkable Inter-Annotator Agreement indicates that the annotation scheme produces consistent annotations of high quality. Subsequent classification experiments using BERT-based models, including ClimateBERT, are presented. These yield encouraging results, while also indicating room for a model specifically tailored for environmental texts. The dataset is made freely available to stimulate further research.

EcoVerse: An Annotated Twitter Dataset for Eco-Relevance Classification, Environmental Impact Analysis, and Stance Detection

lrec-main.485

Poster

2404.05133

https://huggingface.co/papers/2404.05133

https://aclanthology.org/2024.lrec-main.486.bib

https://aclanthology.org/2024.lrec-main.486/

@inproceedings{t-y-s-s-etal-2024-ecthr, title = "{EC}t{HR}-{PCR}: A Dataset for Precedent Understanding and Prior Case Retrieval in the {E}uropean Court of Human Rights", author = "T.y.s.s., Santosh and Haddad, Rashid and Grabmair, Matthias", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.486", pages = "5473--5483", abstract = "In common law jurisdictions, legal practitioners rely on precedents to construct arguments, in line with the doctrine of stare decisis. As the number of cases grow over the years, prior case retrieval (PCR) has garnered significant attention. Besides lacking real-world scale, existing PCR datasets do not simulate a realistic setting, because their queries use complete case documents while only masking references to prior cases. The query is thereby exposed to legal reasoning not yet available when constructing an argument for an undecided case as well as spurious patterns left behind by citation masks, potentially short-circuiting a comprehensive understanding of case facts and legal principles. To address these limitations, we introduce a PCR dataset based on judgements from the European Court of Human Rights (ECtHR), which explicitly separate facts from arguments and exhibit precedential practices, aiding us to develop this PCR dataset to foster systems{'} comprehensive understanding. We benchmark different lexical and dense retrieval approaches with various negative sampling strategies, adapting them to deal with long text sequences using hierarchical variants. We found that difficulty-based negative sampling strategies were not effective for the PCR task, highlighting the need for investigation into domain-specific difficulty criteria. Furthermore, we observe performance of the dense models degrade with time and calls for further research into temporal adaptation of retrieval models. Additionally, we assess the influence of different views , Halsbury{'}s and Goodhart{'}s, in practice in ECtHR jurisdiction using PCR task.", }

In common law jurisdictions, legal practitioners rely on precedents to construct arguments, in line with the doctrine of stare decisis. As the number of cases grow over the years, prior case retrieval (PCR) has garnered significant attention. Besides lacking real-world scale, existing PCR datasets do not simulate a realistic setting, because their queries use complete case documents while only masking references to prior cases. The query is thereby exposed to legal reasoning not yet available when constructing an argument for an undecided case as well as spurious patterns left behind by citation masks, potentially short-circuiting a comprehensive understanding of case facts and legal principles. To address these limitations, we introduce a PCR dataset based on judgements from the European Court of Human Rights (ECtHR), which explicitly separate facts from arguments and exhibit precedential practices, aiding us to develop this PCR dataset to foster systems{'} comprehensive understanding. We benchmark different lexical and dense retrieval approaches with various negative sampling strategies, adapting them to deal with long text sequences using hierarchical variants. We found that difficulty-based negative sampling strategies were not effective for the PCR task, highlighting the need for investigation into domain-specific difficulty criteria. Furthermore, we observe performance of the dense models degrade with time and calls for further research into temporal adaptation of retrieval models. Additionally, we assess the influence of different views , Halsbury{'}s and Goodhart{'}s, in practice in ECtHR jurisdiction using PCR task.

ECtHR-PCR: A Dataset for Precedent Understanding and Prior Case Retrieval in the European Court of Human Rights

lrec-main.486

Poster

2404.00596

https://huggingface.co/papers/2404.00596

https://aclanthology.org/2024.lrec-main.487.bib

https://aclanthology.org/2024.lrec-main.487/

@inproceedings{ding-etal-2024-edda, title = "{EDDA}: An Encoder-Decoder Data Augmentation Framework for Zero-Shot Stance Detection", author = "Ding, Daijun and Dong, Li and Huang, Zhichao and Xu, Guangning and Huang, Xu and Liu, Bo and Jing, Liwen and Zhang, Bowen", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.487", pages = "5484--5494", abstract = "Stance detection aims to determine the attitude expressed in text towards a given target. Zero-shot stance detection (ZSSD) has emerged to classify stances towards unseen targets during inference. Recent data augmentation techniques for ZSSD increase transferable knowledge between targets through text or target augmentation. However, these methods exhibit limitations. Target augmentation lacks logical connections between generated targets and source text, while text augmentation relies solely on training data, resulting in insufficient generalization. To address these issues, we propose an encoder-decoder data augmentation (EDDA) framework. The encoder leverages large language models and chain-of-thought prompting to summarize texts into target-specific if-then rationales, establishing logical relationships. The decoder generates new samples based on these expressions using a semantic correlation word replacement strategy to increase syntactic diversity. We also analyze the generated expressions to develop a rationale-enhanced network that fully utilizes the augmented data. Experiments on benchmark datasets demonstrate our approach substantially improves over state-of-the-art ZSSD techniques. The proposed EDDA framework increases semantic relevance and syntactic variety in augmented texts while enabling interpretable rationale-based learning.", }

Stance detection aims to determine the attitude expressed in text towards a given target. Zero-shot stance detection (ZSSD) has emerged to classify stances towards unseen targets during inference. Recent data augmentation techniques for ZSSD increase transferable knowledge between targets through text or target augmentation. However, these methods exhibit limitations. Target augmentation lacks logical connections between generated targets and source text, while text augmentation relies solely on training data, resulting in insufficient generalization. To address these issues, we propose an encoder-decoder data augmentation (EDDA) framework. The encoder leverages large language models and chain-of-thought prompting to summarize texts into target-specific if-then rationales, establishing logical relationships. The decoder generates new samples based on these expressions using a semantic correlation word replacement strategy to increase syntactic diversity. We also analyze the generated expressions to develop a rationale-enhanced network that fully utilizes the augmented data. Experiments on benchmark datasets demonstrate our approach substantially improves over state-of-the-art ZSSD techniques. The proposed EDDA framework increases semantic relevance and syntactic variety in augmented texts while enabling interpretable rationale-based learning.

EDDA: An Encoder-Decoder Data Augmentation Framework for Zero-Shot Stance Detection

lrec-main.487

Poster

https://aclanthology.org/2024.lrec-main.488.bib

https://aclanthology.org/2024.lrec-main.488/

@inproceedings{touileb-etal-2024-eden, title = "{EDEN}: A Dataset for Event Detection in {N}orwegian News", author = "Touileb, Samia and Murstad, Jeanett and M{\ae}hlum, Petter and Steskal, Lubos and Storset, Lilja Charlotte and You, Huiling and {\O}vrelid, Lilja", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.488", pages = "5495--5506", abstract = "We present EDEN, the first Norwegian dataset annotated with event information at the sentence level, adapting the widely used ACE event schema to Norwegian. The paper describes the manual annotation of Norwegian text as well as transcribed speech in the news domain, together with inter-annotator agreement and discussions of relevant dataset statistics. We also present preliminary modeling results using a graph-based event parser. The resulting dataset will be freely available for download and use.", }

We present EDEN, the first Norwegian dataset annotated with event information at the sentence level, adapting the widely used ACE event schema to Norwegian. The paper describes the manual annotation of Norwegian text as well as transcribed speech in the news domain, together with inter-annotator agreement and discussions of relevant dataset statistics. We also present preliminary modeling results using a graph-based event parser. The resulting dataset will be freely available for download and use.

EDEN: A Dataset for Event Detection in Norwegian News

lrec-main.488

Poster

https://aclanthology.org/2024.lrec-main.489.bib

https://aclanthology.org/2024.lrec-main.489/

@inproceedings{di-nuovo-etal-2024-educational, title = "Educational Dialogue Systems for Visually Impaired Students: Introducing a Task-Oriented User-Agent Corpus", author = "Di Nuovo, Elisa and Sanguinetti, Manuela and Balestrucci, Pier Felice and Anselma, Luca and Bernareggi, Cristian and Mazzei, Alessandro", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.489", pages = "5507--5519", abstract = "This paper describes a corpus consisting of real-world dialogues in English between users and a task-oriented conversational agent, with interactions revolving around the description of finite state automata. The creation of this corpus is part of a larger research project aimed at developing tools for an easier access to educational content, especially in STEM fields, for users with visual impairments. The development of this corpus was precisely motivated by the aim of providing a useful resource to support the design of such tools. The core feature of this corpus is that its creation involved both sighted and visually impaired participants, thus allowing for a greater diversity of perspectives and giving the opportunity to identify possible differences in the way the two groups of participants interacted with the agent. The paper introduces this corpus, giving an account of the process that led to its creation, i.e. the methodology followed to obtain the data, the annotation scheme adopted, and the analysis of the results. Finally, the paper reports the results of a classification experiment on the annotated corpus, and an additional experiment to assess the annotation capabilities of three large language models, in view of a further expansion of the corpus.", }

This paper describes a corpus consisting of real-world dialogues in English between users and a task-oriented conversational agent, with interactions revolving around the description of finite state automata. The creation of this corpus is part of a larger research project aimed at developing tools for an easier access to educational content, especially in STEM fields, for users with visual impairments. The development of this corpus was precisely motivated by the aim of providing a useful resource to support the design of such tools. The core feature of this corpus is that its creation involved both sighted and visually impaired participants, thus allowing for a greater diversity of perspectives and giving the opportunity to identify possible differences in the way the two groups of participants interacted with the agent. The paper introduces this corpus, giving an account of the process that led to its creation, i.e. the methodology followed to obtain the data, the annotation scheme adopted, and the analysis of the results. Finally, the paper reports the results of a classification experiment on the annotated corpus, and an additional experiment to assess the annotation capabilities of three large language models, in view of a further expansion of the corpus.

Educational Dialogue Systems for Visually Impaired Students: Introducing a Task-Oriented User-Agent Corpus

lrec-main.489

Poster

https://aclanthology.org/2024.lrec-main.490.bib

https://aclanthology.org/2024.lrec-main.490/

@inproceedings{hu-etal-2024-eee, title = "{EEE}-{QA}: Exploring Effective and Efficient Question-Answer Representations", author = "Hu, Zhanghao and Yang, Yijun and Xu, Junjie and Qiu, Yifu and Chen, Pinzhen", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.490", pages = "5520--5525", abstract = "Current approaches to question answering rely on pre-trained language models (PLMs) like RoBERTa. This work challenges the existing question-answer encoding convention and explores finer representations. We begin with testing various pooling methods compared to using the begin-of-sentence token as a question representation for better quality. Next, we explore opportunities to simultaneously embed all answer candidates with the question. This enables cross-reference between answer choices and improves inference throughput via reduced memory usage. Despite their simplicity and effectiveness, these methods have yet to be widely studied in current frameworks. We experiment with different PLMs, and with and without the integration of knowledge graphs. Results prove that the memory efficacy of the proposed techniques with little sacrifice in performance. Practically, our work enhances 38-100{\%} throughput with 26-65{\%} speedups on consumer-grade GPUs by allowing for considerably larger batch sizes. Our work sends a message to the community with promising directions in both representation quality and efficiency for the question-answering task in natural language processing.", }

Current approaches to question answering rely on pre-trained language models (PLMs) like RoBERTa. This work challenges the existing question-answer encoding convention and explores finer representations. We begin with testing various pooling methods compared to using the begin-of-sentence token as a question representation for better quality. Next, we explore opportunities to simultaneously embed all answer candidates with the question. This enables cross-reference between answer choices and improves inference throughput via reduced memory usage. Despite their simplicity and effectiveness, these methods have yet to be widely studied in current frameworks. We experiment with different PLMs, and with and without the integration of knowledge graphs. Results prove that the memory efficacy of the proposed techniques with little sacrifice in performance. Practically, our work enhances 38-100{\%} throughput with 26-65{\%} speedups on consumer-grade GPUs by allowing for considerably larger batch sizes. Our work sends a message to the community with promising directions in both representation quality and efficiency for the question-answering task in natural language processing.

EEE-QA: Exploring Effective and Efficient Question-Answer Representations

lrec-main.490

Poster

2403.02176

https://aclanthology.org/2024.lrec-main.491.bib

https://aclanthology.org/2024.lrec-main.491/

@inproceedings{beniamine-etal-2024-eesthetic, title = "Eesthetic: A Paralex Lexicon of {E}stonian Paradigms", author = "Beniamine, Sacha and Aigro, Mari and Baerman, Matthew and Bouton, Jules and Copot, Maria", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.491", pages = "5526--5537", abstract = "We introduce Eesthetic, a comprehensive Estonian noun and verb lexicon sourced from the Ekilex database. It documents 5475 nouns inflecting for 28 paradigm cells and 5076 verbs inflecting for 51 cells, and comprises a total of 452885 inflected forms. Our openly accessible machine-readable dataset adheres to the Paralex standard. It comprises CSV tables linked by formal relationships. Metadata in JSON format, following the Frictionless standard, provides detailed descriptions of the tables and dataset. The lexicon offers extensive linguistic annotations, including orthographic forms, automatically transcribed phonemic transcriptions, non-canonical morphological phenomena such as overabundance and defectiveness, rich mapping of the paradigm cells and feature-values to other notation schemes, a decomposition of phonemes in distinctive features, and annotation of inflection classes. It is suited for both monolingual and comparative research, enabling qualitative and quantitative analysis. This paper outlines the creation process, rationale, and resulting structure, along with our set of rules for automatic orthography-to-phonemic transcription conversion.", }

We introduce Eesthetic, a comprehensive Estonian noun and verb lexicon sourced from the Ekilex database. It documents 5475 nouns inflecting for 28 paradigm cells and 5076 verbs inflecting for 51 cells, and comprises a total of 452885 inflected forms. Our openly accessible machine-readable dataset adheres to the Paralex standard. It comprises CSV tables linked by formal relationships. Metadata in JSON format, following the Frictionless standard, provides detailed descriptions of the tables and dataset. The lexicon offers extensive linguistic annotations, including orthographic forms, automatically transcribed phonemic transcriptions, non-canonical morphological phenomena such as overabundance and defectiveness, rich mapping of the paradigm cells and feature-values to other notation schemes, a decomposition of phonemes in distinctive features, and annotation of inflection classes. It is suited for both monolingual and comparative research, enabling qualitative and quantitative analysis. This paper outlines the creation process, rationale, and resulting structure, along with our set of rules for automatic orthography-to-phonemic transcription conversion.

Eesthetic: A Paralex Lexicon of Estonian Paradigms

lrec-main.491

Poster

https://aclanthology.org/2024.lrec-main.492.bib

https://aclanthology.org/2024.lrec-main.492/

@inproceedings{yang-etal-2024-effective, title = "Effective Distillation of Table-based Reasoning Ability from {LLM}s", author = "Yang, Bohao and Tang, Chen and Zhao, Kun and Xiao, Chenghao and Lin, Chenghua", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.492", pages = "5538--5550", abstract = "Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, their enormous parameter size and extremely high requirements for compute power pose challenges for their practical deployment. Recent research has revealed that specific capabilities of LLMs, such as numerical reasoning, can be transferred to smaller models through distillation. Some studies explore the potential of leveraging LLMs to perform table-based reasoning. However, there has been no prior work focusing on table reasoning skills in smaller models specifically tailored for scientific table-to-text generation tasks. In this paper, we propose a novel table-based reasoning distillation approach, with the aim of distilling LLMs into tailored smaller models. Our experimental results have shown that a 220 million parameter model (Flan-T5-base) fine-tuned using distilled data, not only achieves a significant improvement compared to traditionally fine-tuned baselines, but also surpasses specific LLMs on a scientific table-to-text generation dataset. Our code is available at https://github.com/Bernard-Yang/DistillTableCoT.", }

Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, their enormous parameter size and extremely high requirements for compute power pose challenges for their practical deployment. Recent research has revealed that specific capabilities of LLMs, such as numerical reasoning, can be transferred to smaller models through distillation. Some studies explore the potential of leveraging LLMs to perform table-based reasoning. However, there has been no prior work focusing on table reasoning skills in smaller models specifically tailored for scientific table-to-text generation tasks. In this paper, we propose a novel table-based reasoning distillation approach, with the aim of distilling LLMs into tailored smaller models. Our experimental results have shown that a 220 million parameter model (Flan-T5-base) fine-tuned using distilled data, not only achieves a significant improvement compared to traditionally fine-tuned baselines, but also surpasses specific LLMs on a scientific table-to-text generation dataset. Our code is available at https://github.com/Bernard-Yang/DistillTableCoT.

Effective Distillation of Table-based Reasoning Ability from LLMs

lrec-main.492

Poster

2309.13182

https://huggingface.co/papers/2309.13182

https://aclanthology.org/2024.lrec-main.493.bib

https://aclanthology.org/2024.lrec-main.493/

@inproceedings{ou-jian-2024-effective, title = "Effective Integration of Text Diffusion and Pre-Trained Language Models with Linguistic Easy-First Schedule", author = "Ou, Yimin and Jian, Ping", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.493", pages = "5551--5561", abstract = "Diffusion models have become a powerful generative modeling paradigm, achieving great success in continuous data patterns. However, the discrete nature of text data results in compatibility issues between continuous diffusion models (CDMs) and pre-trained language models (PLMs). That is, the performance of diffusion models even degrades when combined with PLMs. To alleviate this issue, we propose to utilize a pre-trained decoder to convert the denoised embedding vectors into natural language instead of using the widely used rounding operation. In this way, CDMs can be more effectively combined with PLMs. Additionally, considering that existing noise schedules in text diffusion models do not take into account the linguistic differences among tokens, which violates the easy-first policy for text generation, we propose a linguistic easy-first schedule that incorporates the measure of word importance, conforming to easy-first-generation linguistic features and bringing about improved generation quality. Experiment results on the E2E dataset and five controllable tasks show that our approach can combine the merits of CDMs and PLMs, significantly outperforming other diffusion-based models.", }

Diffusion models have become a powerful generative modeling paradigm, achieving great success in continuous data patterns. However, the discrete nature of text data results in compatibility issues between continuous diffusion models (CDMs) and pre-trained language models (PLMs). That is, the performance of diffusion models even degrades when combined with PLMs. To alleviate this issue, we propose to utilize a pre-trained decoder to convert the denoised embedding vectors into natural language instead of using the widely used rounding operation. In this way, CDMs can be more effectively combined with PLMs. Additionally, considering that existing noise schedules in text diffusion models do not take into account the linguistic differences among tokens, which violates the easy-first policy for text generation, we propose a linguistic easy-first schedule that incorporates the measure of word importance, conforming to easy-first-generation linguistic features and bringing about improved generation quality. Experiment results on the E2E dataset and five controllable tasks show that our approach can combine the merits of CDMs and PLMs, significantly outperforming other diffusion-based models.

Effective Integration of Text Diffusion and Pre-Trained Language Models with Linguistic Easy-First Schedule

lrec-main.493

Poster

https://aclanthology.org/2024.lrec-main.494.bib

https://aclanthology.org/2024.lrec-main.494/

@inproceedings{yee-etal-2024-efficiency, title = "Efficiency and Effectiveness in Task-Oriented Dialogue: On Construction Repetition, Information Rate, and Task Success", author = "Yee, Jun Sen and Giulianelli, Mario and Sinclair, Arabella J.", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.494", pages = "5562--5577", abstract = "We investigate the roles that efficiency and effectiveness play in speakers{'} repetition of shared word sequences, or constructions, in task-oriented dialogue. We find that repeating constructions has negative effects on information rate and positive effects on rate of delivery, that information rate managing strategies are predictive of task success, and that this varies by the communicative function of the constructions being repeated. More effective dialogue is characterised by greater levels of shared construction usage and more efficient task-related repetition; while task-agnostic repetition can seem redundant, it can serve important efficiency and effectiveness functions. Our results provide a nuanced picture of the importance of repetition and of developing a shared lexicon for both efficiency and effectiveness in task-oriented dialogue.", }

We investigate the roles that efficiency and effectiveness play in speakers{'} repetition of shared word sequences, or constructions, in task-oriented dialogue. We find that repeating constructions has negative effects on information rate and positive effects on rate of delivery, that information rate managing strategies are predictive of task success, and that this varies by the communicative function of the constructions being repeated. More effective dialogue is characterised by greater levels of shared construction usage and more efficient task-related repetition; while task-agnostic repetition can seem redundant, it can serve important efficiency and effectiveness functions. Our results provide a nuanced picture of the importance of repetition and of developing a shared lexicon for both efficiency and effectiveness in task-oriented dialogue.

Efficiency and Effectiveness in Task-Oriented Dialogue: On Construction Repetition, Information Rate, and Task Success

lrec-main.494

Poster

https://aclanthology.org/2024.lrec-main.495.bib

https://aclanthology.org/2024.lrec-main.495/

@inproceedings{martinez-lorenzo-navigli-2024-efficient, title = "Efficient {AMR} Parsing with {CLAP}: Compact Linearization with an Adaptable Parser", author = "Martinez Lorenzo, Abelardo Carlos and Navigli, Roberto", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.495", pages = "5578--5584", abstract = "Sequence-to-sequence models have become the de facto standard for Abstract Meaning Representation (AMR) parsing due to their high-quality performance. However, these systems face efficiency challenges because of their large model size and computational time, which limit their accessibility within the research community. This paper aims to break down these barriers by introducing a novel linearization and system that significantly enhances the efficiency and accessibility of previous AMR parsers. First, we propose our novel Compact linearization that simplifies encoding, thereby reducing the number of tokens by between 40{\%} and 50{\%}. Second, we present CLAP, an innovative modular system that maintains the model{'}s high performance while achieving remarkable 80{\%} reduction in training and inference times. Furthermore, CLAP is compatible with multiple autoregressive Language Models (LM) and tokenizers, such as BART, T5, and others. These advancements underscore the importance of optimizing sequence-to-sequence models in AMR parsing, thus democratizing access to high-quality semantic analysis. Our code is publicly available at https://github.com/SapienzaNLP/clap/.", }

Sequence-to-sequence models have become the de facto standard for Abstract Meaning Representation (AMR) parsing due to their high-quality performance. However, these systems face efficiency challenges because of their large model size and computational time, which limit their accessibility within the research community. This paper aims to break down these barriers by introducing a novel linearization and system that significantly enhances the efficiency and accessibility of previous AMR parsers. First, we propose our novel Compact linearization that simplifies encoding, thereby reducing the number of tokens by between 40{\%} and 50{\%}. Second, we present CLAP, an innovative modular system that maintains the model{'}s high performance while achieving remarkable 80{\%} reduction in training and inference times. Furthermore, CLAP is compatible with multiple autoregressive Language Models (LM) and tokenizers, such as BART, T5, and others. These advancements underscore the importance of optimizing sequence-to-sequence models in AMR parsing, thus democratizing access to high-quality semantic analysis. Our code is publicly available at https://github.com/SapienzaNLP/clap/.

Efficient AMR Parsing with CLAP: Compact Linearization with an Adaptable Parser

lrec-main.495

Poster

https://aclanthology.org/2024.lrec-main.496.bib

https://aclanthology.org/2024.lrec-main.496/

@inproceedings{sun-etal-2024-efficient, title = "Efficient and Accurate Contextual Re-Ranking for Knowledge Graph Question Answering", author = "Sun, Kexuan and Jedema, Nicolaas Paul and Sharma, Karishma and Janssen, Ruben and Pujara, Jay and Szekely, Pedro and Moschitti, Alessandro", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.496", pages = "5585--5595", abstract = "The efficacy of neural {``}retrieve and generate{''} systems is well established for question answering (QA) over unstructured text. Recent efforts seek to extend this approach to knowledge graph (KG) QA by converting structured triples to unstructured text. However, the relevance of KG triples retrieved by these systems limits their accuracy. In this paper, we improve the relevance of retrieved triples using a carefully designed re-ranker. Specifically, our pipeline (i) retrieves over documents of triples grouped by entity, (ii) re-ranks triples from these documents with context: triples in the 1-hop neighborhood of the documents{'} subject entity, and (iii) generates an answer from highly relevant re-ranked triples. To train our re-ranker, we propose a novel {``}triple-level{''} labeling strategy that infers fine-grained labels and shows that these significantly improve the relevance of retrieved information. We show that the resulting {``}retrieve, re-rank, and generate{''} pipeline significantly improves upon prior KGQA systems, achieving a new state-of-the-art on FreebaseQA by 5.56{\%} Exact Match. We perform multiple ablations that reveal the distinct benefits of our contextual re-ranker and labeling strategy and conclude with a case study that highlights opportunities for future works.", }

The efficacy of neural {``}retrieve and generate{''} systems is well established for question answering (QA) over unstructured text. Recent efforts seek to extend this approach to knowledge graph (KG) QA by converting structured triples to unstructured text. However, the relevance of KG triples retrieved by these systems limits their accuracy. In this paper, we improve the relevance of retrieved triples using a carefully designed re-ranker. Specifically, our pipeline (i) retrieves over documents of triples grouped by entity, (ii) re-ranks triples from these documents with context: triples in the 1-hop neighborhood of the documents{'} subject entity, and (iii) generates an answer from highly relevant re-ranked triples. To train our re-ranker, we propose a novel {``}triple-level{''} labeling strategy that infers fine-grained labels and shows that these significantly improve the relevance of retrieved information. We show that the resulting {``}retrieve, re-rank, and generate{''} pipeline significantly improves upon prior KGQA systems, achieving a new state-of-the-art on FreebaseQA by 5.56{\%} Exact Match. We perform multiple ablations that reveal the distinct benefits of our contextual re-ranker and labeling strategy and conclude with a case study that highlights opportunities for future works.

Efficient and Accurate Contextual Re-Ranking for Knowledge Graph Question Answering

lrec-main.496

Poster

https://aclanthology.org/2024.lrec-main.497.bib

https://aclanthology.org/2024.lrec-main.497/

@inproceedings{munoz-etal-2024-eftnas, title = "{EFTNAS}: Searching for Efficient Language Models in First-Order Weight-Reordered Super-Networks", author = "Munoz, Juan Pablo and Zheng, Yi and Jain, Nilesh", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.497", pages = "5596--5608", abstract = "Transformer-based models have demonstrated outstanding performance in natural language processing (NLP) tasks and many other domains, e.g., computer vision. Depending on the size of these models, which have grown exponentially in the past few years, machine learning practitioners might be restricted from deploying them in resource-constrained environments. This paper discusses the compression of transformer-based models for multiple resource budgets. Integrating neural architecture search (NAS) and network pruning techniques, we effectively generate and train weight-sharing super-networks that contain efficient, high-performing, and compressed transformer-based models. A common challenge in NAS is the design of the search space, for which we propose a method to automatically obtain the boundaries of the search space and then derive the rest of the intermediate possible architectures using a first-order weight importance technique. The proposed end-to-end NAS solution, EFTNAS, discovers efficient subnetworks that have been compressed and fine-tuned for downstream NLP tasks. We demonstrate EFTNAS on the General Language Understanding Evaluation (GLUE) benchmark and the Stanford Question Answering Dataset (SQuAD), obtaining high-performing smaller models with a reduction of more than 5x in size without or with little degradation in performance.", }

Transformer-based models have demonstrated outstanding performance in natural language processing (NLP) tasks and many other domains, e.g., computer vision. Depending on the size of these models, which have grown exponentially in the past few years, machine learning practitioners might be restricted from deploying them in resource-constrained environments. This paper discusses the compression of transformer-based models for multiple resource budgets. Integrating neural architecture search (NAS) and network pruning techniques, we effectively generate and train weight-sharing super-networks that contain efficient, high-performing, and compressed transformer-based models. A common challenge in NAS is the design of the search space, for which we propose a method to automatically obtain the boundaries of the search space and then derive the rest of the intermediate possible architectures using a first-order weight importance technique. The proposed end-to-end NAS solution, EFTNAS, discovers efficient subnetworks that have been compressed and fine-tuned for downstream NLP tasks. We demonstrate EFTNAS on the General Language Understanding Evaluation (GLUE) benchmark and the Stanford Question Answering Dataset (SQuAD), obtaining high-performing smaller models with a reduction of more than 5x in size without or with little degradation in performance.

EFTNAS: Searching for Efficient Language Models in First-Order Weight-Reordered Super-Networks

lrec-main.497

Poster

https://aclanthology.org/2024.lrec-main.498.bib

https://aclanthology.org/2024.lrec-main.498/

@inproceedings{sun-etal-2024-eliciting, title = "Eliciting Motivational Interviewing Skill Codes in Psychotherapy with {LLM}s: A Bilingual Dataset and Analytical Study", author = "Sun, Xin and Pei, Jiahuan and Wit, Jan de and Aliannejadi, Mohammad and Krahmer, Emiel and Dobber, Jos T.P. and Bosch, Jos A.", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.498", pages = "5609--5621", abstract = "Behavioral coding (BC) in motivational interviewing (MI) holds great potential for enhancing the efficacy of MI counseling. However, manual coding is labor-intensive, and automation efforts are hindered by the lack of data due to the privacy of psychotherapy. To address these challenges, we introduce BiMISC, a bilingual dataset of MI conversations in English and Dutch, sourced from real counseling sessions. Expert annotations in BiMISC adhere strictly to the motivational interviewing skills code (MISC) scheme, offering a pivotal resource for MI research. Additionally, we present a novel approach to elicit the MISC expertise from Large language models (LLMs) for MI coding. Through the in-depth analysis of BiMISC and the evaluation of our proposed approach, we demonstrate that the LLM-based approach yields results closely aligned with expert annotations and maintains consistent performance across different languages. Our contributions not only furnish the MI community with a valuable bilingual dataset but also spotlight the potential of LLMs in MI coding, laying the foundation for future MI research.", }

Behavioral coding (BC) in motivational interviewing (MI) holds great potential for enhancing the efficacy of MI counseling. However, manual coding is labor-intensive, and automation efforts are hindered by the lack of data due to the privacy of psychotherapy. To address these challenges, we introduce BiMISC, a bilingual dataset of MI conversations in English and Dutch, sourced from real counseling sessions. Expert annotations in BiMISC adhere strictly to the motivational interviewing skills code (MISC) scheme, offering a pivotal resource for MI research. Additionally, we present a novel approach to elicit the MISC expertise from Large language models (LLMs) for MI coding. Through the in-depth analysis of BiMISC and the evaluation of our proposed approach, we demonstrate that the LLM-based approach yields results closely aligned with expert annotations and maintains consistent performance across different languages. Our contributions not only furnish the MI community with a valuable bilingual dataset but also spotlight the potential of LLMs in MI coding, laying the foundation for future MI research.

Eliciting Motivational Interviewing Skill Codes in Psychotherapy with LLMs: A Bilingual Dataset and Analytical Study

lrec-main.498

Poster

https://aclanthology.org/2024.lrec-main.499.bib

https://aclanthology.org/2024.lrec-main.499/

@inproceedings{riaz-etal-2024-ellen, title = "{ELLEN}: Extremely Lightly Supervised Learning for Efficient Named Entity Recognition", author = "Riaz, Haris and Dumitru, Razvan Gabriel and Surdeanu, Mihai", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.499", pages = "5622--5636", abstract = "In this work, we revisit the problem of semi-supervised named entity recognition (NER) focusing on extremely light supervision, consisting of a lexicon containing only 10 examples per class. We introduce ELLEN, a simple, fully modular, neuro-symbolic method that blends fine-tuned language models with linguistic rules. These rules include insights such as {``}One Sense Per Discourse{''}, using a Masked Language Model as an unsupervised NER, leveraging part-of-speech tags to identify and eliminate unlabeled entities as false negatives, and other intuitions about classifier confidence scores in local and global context. ELLEN achieves very strong performance on the CoNLL-2003 dataset when using the minimal supervision from the lexicon above. It also outperforms most existing (and considerably more complex) semi-supervised NER methods under the same supervision settings commonly used in the literature (i.e., 5{\%} of the training data). Further, we evaluate our CoNLL-2003 model in a zero-shot scenario on WNUT-17 where we find that it outperforms GPT-3.5 and achieves comparable performance to GPT-4. In a zero-shot setting, ELLEN also achieves over 75{\%} of the performance of a strong, fully supervised model trained on gold data. Our code is publicly available.", }

In this work, we revisit the problem of semi-supervised named entity recognition (NER) focusing on extremely light supervision, consisting of a lexicon containing only 10 examples per class. We introduce ELLEN, a simple, fully modular, neuro-symbolic method that blends fine-tuned language models with linguistic rules. These rules include insights such as {``}One Sense Per Discourse{''}, using a Masked Language Model as an unsupervised NER, leveraging part-of-speech tags to identify and eliminate unlabeled entities as false negatives, and other intuitions about classifier confidence scores in local and global context. ELLEN achieves very strong performance on the CoNLL-2003 dataset when using the minimal supervision from the lexicon above. It also outperforms most existing (and considerably more complex) semi-supervised NER methods under the same supervision settings commonly used in the literature (i.e., 5{\%} of the training data). Further, we evaluate our CoNLL-2003 model in a zero-shot scenario on WNUT-17 where we find that it outperforms GPT-3.5 and achieves comparable performance to GPT-4. In a zero-shot setting, ELLEN also achieves over 75{\%} of the performance of a strong, fully supervised model trained on gold data. Our code is publicly available.

ELLEN: Extremely Lightly Supervised Learning for Efficient Named Entity Recognition

lrec-main.499

Poster

2403.17385

https://aclanthology.org/2024.lrec-main.500.bib

https://aclanthology.org/2024.lrec-main.500/

@inproceedings{kallas-etal-2024-emad, title = "{EMAD}: A Bridge Tagset for Unifying {A}rabic {POS} Annotations", author = "Kallas, Omar and Inoue, Go and Habash, Nizar", editor = "Calzolari, Nicoletta and Kan, Min-Yen and Hoste, Veronique and Lenci, Alessandro and Sakti, Sakriani and Xue, Nianwen", booktitle = "Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)", month = may, year = "2024", address = "Torino, Italia", publisher = "ELRA and ICCL", url = "https://aclanthology.org/2024.lrec-main.500", pages = "5637--5643", abstract = "There have been many attempts to model the morphological richness and complexity of Arabic, leading to numerous Part-of-Speech (POS) tagsets that differ in terms of (a) which morphological features they represent, (b) how they represent them, and (c) the degree of specification of said features. Tagset granularity plays an important role in determining how annotated data can be used and for what applications. Due to the diversity among existing tagsets, many annotated corpora for Arabic cannot be easily combined, which exacerbates the Arabic resource poverty situation. In this work, we propose an intermediate tagset designed to facilitate the conversion and unification of different tagsets used to annotate Arabic corpora. This new tagset acts as a bridge between different annotation schemes, simplifying the integration of annotated corpora and promoting collaboration across the projects using them.", }

There have been many attempts to model the morphological richness and complexity of Arabic, leading to numerous Part-of-Speech (POS) tagsets that differ in terms of (a) which morphological features they represent, (b) how they represent them, and (c) the degree of specification of said features. Tagset granularity plays an important role in determining how annotated data can be used and for what applications. Due to the diversity among existing tagsets, many annotated corpora for Arabic cannot be easily combined, which exacerbates the Arabic resource poverty situation. In this work, we propose an intermediate tagset designed to facilitate the conversion and unification of different tagsets used to annotate Arabic corpora. This new tagset acts as a bridge between different annotation schemes, simplifying the integration of annotated corpora and promoting collaboration across the projects using them.

EMAD: A Bridge Tagset for Unifying Arabic POS Annotations

lrec-main.500

Poster