Datasets:

allenai
/

SciRIFF

input stringlengths 286 19k	output stringlengths 1 15.8k	metadata dict	_instance_id stringlengths 15 62
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Thus, over the past few years, along with advances in the use of learning and statistical methods for acquisition of full parsers (Collins, 1997; Charniak, 1997a; Charniak, 1997b; Ratnaparkhi, 1997), significant progress has been made on the use of statistical learning methods to recognize shallow parsing patterns syntactic phrases or words that participate in a syntactic relationship (Church, 1988; Ramshaw and Marcus, 1995; Argamon et al., 1998; Cardie and Pierce, 1998; Munoz et al., 1999; Punyakanok and Roth, 2001; Buchholz et al., 1999; Tjong Kim Sang and Buchholz, 2000). While earlier work in this direction concentrated on manual construction of rules, most of the recent work has been motivated by the observation that shallow syntactic information can be extracted using local information by examining the pattern itself, its nearby context and the local part-of-speech information. to ] [NP only $ 1.8 billion ] [PP in ] [NP September] . Citation Sentence: Thus , over the past few years , along with advances in the use of learning and statistical methods for acquisition of full parsers ( Collins , 1997 ; Charniak , 1997a ; Charniak , 1997b ; Ratnaparkhi , 1997 ) , significant progress has been made on the use of statistical learning methods to recognize shallow parsing patterns syntactic phrases or words that participate in a syntactic relationship ( Church , 1988 ; Ramshaw and Marcus , 1995 ; Argamon et al. , 1998 ; Cardie and Pierce , 1998 ; Munoz et al. , 1999 ; Punyakanok and Roth , 2001 ; Buchholz et al. , 1999 ; Tjong Kim Sang and Buchholz , 2000 ) . Context after the citation: Research on shallow parsing was inspired by psycholinguistics arguments (Gee and Grosjean, 1983) that suggest that in many scenarios (e.g., conversational) full parsing is not a realistic strategy for sentence processing and analysis, and was further motivated by several arguments from a natural language engineering viewpoint. First, it has been noted that in many natural language applications it is sufficient to use shallow parsing information; information such as noun phrases (NPs) and other syntactic sequences have been found useful in many large-scale language processing applications including information extraction and text summarization (Grishman, 1995; Appelt et al., 1993). Second, while training a full parser requires a collection of fully parsed sentences as training corpus, it is possible to train a shallow parser incrementally. If all that is available is a collection of sentences annotated for NPs, it can be used to produce this level of analysis.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:0
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: This was done by MERT optimization (Och, 2003) towards post-edits under the TER target metric. PE-dev was used to obtain w* to define the utility model. PE-test was held out for testing the algorithmsâ progress on unseen data. Citation Sentence: This was done by MERT optimization ( Och , 2003 ) towards post-edits under the TER target metric . Context after the citation: Note that the goal of our experi- ments is not to improve SMT performance over any algorithm that has access to full information to compute w*. Rather, we want to show that learning from weak feedback leads to convergence in regret with respect to the optimal model, albeit at a slower rate than learning from strong feedback. The feedback data in this experiment were generated by searching the n-best list for translations that are Î±-informative at Î± E 10.1, 0.5, 1.01 (with possible non-zero slack).	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: Similar to our previous work (Chan and Ng, 2005b), we used the supervised WSD approach described in (Lee and Ng, 2002) for our experiments, using the naive Bayes algorithm as our classifier. Citation Sentence: Similar to our previous work ( Chan and Ng , 2005b ) , we used the supervised WSD approach described in ( Lee and Ng , 2002 ) for our experiments , using the naive Bayes algorithm as our classifier . Context after the citation: Knowledge sources used include partsof-speech, surrounding words, and local collocations. This approach achieves state-of-the-art accuracy. All accuracies reported in our experiments are micro-averages over all test examples. In (Chan and Ng, 2005b), we used a multiclass naive Bayes classifier (denoted by NB) for each word.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:10
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: conclusion Context before the citation: As (Daelemans et al., 1999) show, lexical information improves on NP and VP chunking as well. It is not aimed at handling dependencies, which require heavy use of lexical information (Hindle and Rooth, 1993, for PP attachment). The presented method concerns primarily with phrases, which can be represented by a tree structure. Citation Sentence: As ( Daelemans et al. , 1999 ) show , lexical information improves on NP and VP chunking as well . Context after the citation: Since our method uses raw data, representing lexical entries will require a lot of memory. In a future work, we plan to use the system for providing instance candidates, and disambiguate them using an algorithm more suitable for handling lexical information. An additional possibility is to use word-types, such as a special tag for be-verbs, or for prepositions like 'of' which attaches mainly to nouns (Sekine and Grishman, 1995). In a similar vain to Skut and Brants (1998) and Buchholz et al. (1999), the method extends an existing flat shallow-parsing method to handle composite structures.	FutureWork	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:100
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Some previous works (Bannard and Callison-Burch, 2005; Zhao et al., 2009; Kouylekov et al., 2009) indicate, as main limitations of the mentioned resources, their limited coverage, their low precision, and the fact that they are mostly suitable to capture relations mainly between single words. As emerges from the ablation tests reported in (Bentivogli et al., 2010), even the most common resources proved to have a positive impact on some systems and a negative impact on others. Despite the consensus on the usefulness of lexical knowledge for textual inference, determining the actual impact of these resources is not straightforward, as they always represent one component in complex architectures that may use them in different ways. Citation Sentence: Some previous works ( Bannard and Callison-Burch , 2005 ; Zhao et al. , 2009 ; Kouylekov et al. , 2009 ) indicate , as main limitations of the mentioned resources , their limited coverage , their low precision , and the fact that they are mostly suitable to capture relations mainly between single words . Context after the citation: Addressing CLTE we have to face additional and more problematic issues related to: i) the stronger need of lexical knowledge, and ii) the limited availability of multilingual lexical resources. As regards the first issue, itâs worth noting that in the monolingual scenario simple âbag of wordsâ (or âbag of ngramsâ) approaches are per se sufficient to achieve results above baseline. In contrast, their application in the cross-lingual setting is not a viable solution due to the impossibility to perform direct lexical matches between texts and hypotheses in different languages. This situation makes the availability of multilingual lexical knowledge a necessary condition to bridge the language gap.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1000
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Nugget F-score has been employed as a metric in the TREC question-answering track since 2003, to evaluate so-called definition and âotherâ questions (Voorhees 2003). Recently, there is a growing consensus that an evaluation methodology based on the notion of âinformation nuggetsâ may provide an appropriate framework for assessing the quality of answers to complex questions. In Sections 9 and 10, we have discussed many of these issues. Citation Sentence: Nugget F-score has been employed as a metric in the TREC question-answering track since 2003 , to evaluate so-called definition and `` other '' questions ( Voorhees 2003 ) . Context after the citation: A number of studies (e.g., Hildebrandt, Katz, and Lin 2004) have pointed out shortcomings of the original nugget scoring model, although a number of these issues have been recently addressed (Lin and Demner-Fushman 2005a, 2006b). However, adaptation of the nugget evaluation methodology to a domain as specific as clinical medicine is an endeavor that has yet to be undertaken.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1001
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Manually defined heuristics are used to automatically annotate each tree in the treebank with partially specified HPSG derivation trees: Head/argument/modifier distinctions are made for each node in the tree based on Magerman (1994) and Collins (1997); Miyao, Ninomiya, and Tsujii (2004) and Nakanishi, Miyao, and Tsujii (2004) describe a methodology for acquiring an English HPSG from the Penn-II Treebank. Unlike our approach, those of Xia (1999) and Hockenmaier, Bierner, and Baldridge (2004) include a substantial initial correction and clean-up of the Penn-II trees. Citation Sentence: Manually defined heuristics are used to automatically annotate each tree in the treebank with partially specified HPSG derivation trees : Head/argument/modifier distinctions are made for each node in the tree based on Magerman ( 1994 ) and Collins ( 1997 ) ; Context after the citation: the whole tree is then converted to a binary tree; heuristics are applied to deal with phenomena such as LDDs and coordination and to correct some errors in the treebank, and finally an HPSG category is assigned to each node in the tree in accordance with its CFG category. In the next phase of the process (externalization), HPSG lexical entries are automatically extracted from the annotated trees through the application of âinverse schemata.â	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1002
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: conclusion Context before the citation: Regarding future work, there are many research line that may be followed: i) Capturing more features by employing external knowledge such as ontological, lexical resource or WordNet-based features (Basili et al., 2005a; Basili et al., 2005b; Bloehdorn et al., 2006; Bloehdorn and Moschitti, 2007) or shallow semantic trees, (Giuglea and Moschitti, 2004; Giuglea and Moschitti, 2006; Moschitti and Bejan, 2004; Moschitti et al., 2007; Moschitti, 2008; Moschitti et al., 2008). Our novel composite kernels, which account for the two syntactic structures, are experimented with the appropriate convolution kernels and show significant improvement with respect to the state-ofthe-art in RE. For the design of automatic relation classifiers, we have investigated the impact of dependency structures to the RE task. Citation Sentence: Regarding future work , there are many research line that may be followed : i ) Capturing more features by employing external knowledge such as ontological , lexical resource or WordNet-based features ( Basili et al. , 2005a ; Basili et al. , 2005b ; Bloehdorn et al. , 2006 ; Bloehdorn and Moschitti , 2007 ) or shallow semantic trees , ( Giuglea and Moschitti , 2004 ; Giuglea and Moschitti , 2006 ; Moschitti and Bejan , 2004 ; Moschitti et al. , 2007 ; Moschitti , 2008 ; Moschitti et al. , 2008 ) . Context after the citation: ii) Design a new tree-based structures, which combines the information of both constituent and dependency parses. From dependency trees we can extract more precise but also more sparse relationships (which may cause overfit). From constituent trees, we can extract subtrees constituted by non-terminal symbols (grammar symbols), which provide a better generalization (with a risk of underfitting). iii) Design a new kernel which can integrate the advantages of the constituent and dependency tree.	FutureWork	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1003
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: In addition, there has been much work on the application of linguistic and semantic knowledge to information retrieval; see Lin and Demner-Fushman (2006a) for a brief overview. For an overview of systems designed to answer open-domain factoid questions, the TREC QA track overview papers are a good place to start (Voorhees and Tice 1999). In this section, however, we will attempt to draw connections to other clinical information systems (although not necessarily for question answering) and related domain-specific question-answering systems. Citation Sentence: In addition , there has been much work on the application of linguistic and semantic knowledge to information retrieval ; see Lin and Demner-Fushman ( 2006a ) for a brief overview . Context after the citation: The idea that clinical information systems should be sensitive to the practice of evidence-based medicine is not new. Based on analyses of 4,000 MEDLINE citations, MendoncÂ¸a and Cimino (2001) have studied MeSH terms associated with the four basic clinical tasks of therapy, diagnosis, prognosis, and etiology. The goal was to automatically classify citations for task-specific retrieval, similar in spirit to the Hedges Project (Haynes et al. 1994; Wilczynski, McKibbon, and Haynes 2001). Cimino and MendoncÂ¸a reported good performance for etiology, diagnosis, and in particular therapy, but not prognosis.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1004
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: However, Dunning (1993) claims that the log-likelihood chisquared statistic (G2) is more appropriate for corpus-based NLP. The usual statistic encountered in text books is the Pearson chisquared statistic, denoted X2. There is a choice of which statistic to use in conjunction with the test. Citation Sentence: However , Dunning ( 1993 ) claims that the log-likelihood chisquared statistic ( G2 ) is more appropriate for corpus-based NLP . Context after the citation: In Section 6, we compare the two statistics in a task-based evaluation. For Table 2, the value of G2 is 3.8 and the value of X2 is 2.5. Assuming a level of significance of a = 0.05, the critical value is 12.6 (for 6 degrees of freedom). Thus, for this a value, the null hypothesis would not be rejected for either statistic, and the conclusion would be that there is no reason to suppose p(runl (canine), subj) is not a reasonable approximation of p(rtml (dog), subj).	Motivation	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1005
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: In a log-linear parameterization, for example, a prior that penalizes feature strengths far from 1 can be used to do feature selection and avoid overfitting (Chen and Rosenfeld, 1999). Maximum-posterior estimation tries to maximize P(0)Â·Hi fÎ¸(xi, yi) where P(0) is a prior probability. Maximum-likelihood estimation guesses 0Ë to be the 0 maximizing Hi fÎ¸(xi, yi). Citation Sentence: In a log-linear parameterization , for example , a prior that penalizes feature strengths far from 1 can be used to do feature selection and avoid overfitting ( Chen and Rosenfeld , 1999 ) . Context after the citation: The EM algorithm (Dempster et al., 1977) can maximize these functions. Roughly, the E step guesses hidden information: if (xi, yi) was generated from the current fÎ¸, which FST paths stand a chance of having been the path used? (Guessing the path also guesses the exact input and output.) The M step updates 0 to make those paths more likely.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1006
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: This alignment is done on the basis of both length (Gale and Church [7]) and a notion of cognateness (Simard [16]). After identification of word and sentence boundaries the text is processed into a bi-text by an alignment program. An aligner. Citation Sentence: This alignment is done on the basis of both length ( Gale and Church [ 7 ] ) and a notion of cognateness ( Simard [ 16 ] ) . Context after the citation: 2. Transducers. In order to compare numerical expressions, which often diverge in format between given pairs of languages, normalisation toward a common format is required. This is done with transducers (Kaplan and Kay, [10]).	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1007
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: Since earlier versions of the SNoW based CSCL were used only to identify single phrases (Punyakanok and Roth, 2001; Munoz et al., 1999) and never to identify a collection of several phrases at the same time, as we do here, we also trained and tested it under the exact conditions of CoNLL-2000 (Tjong Kim Sang and Buchholz, 2000) to compare it to other shallow parsers. The outcomes of these classifiers are then combined in a way that satisfies some constraints â non-overlapping constraints in this case â using an efficient constraint satisfaction mechanism that makes use of the confidence in the classifierâs outcomes. Indeed, in CSCL (constraint satisfaction with classifiers), SNoW is used to learn several different classifiers â each detects the beginning or end of a phrase of some type (noun phrase, verb phrase, etc.). Citation Sentence: Since earlier versions of the SNoW based CSCL were used only to identify single phrases ( Punyakanok and Roth , 2001 ; Munoz et al. , 1999 ) and never to identify a collection of several phrases at the same time , as we do here , we also trained and tested it under the exact conditions of CoNLL-2000 ( Tjong Kim Sang and Buchholz , 2000 ) to compare it to other shallow parsers . Context after the citation: Table 1 shows that it ranks among the top shallow parsers evaluated there 1.	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1008
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: A more detailed discussion of the various available Arabic tag sets can be found in Habash (2010). These tag sets are hybrids in the sense that they are neither simply the core POS, nor the complete morphologically enriched tag set, but instead they selectively enrich the core POS tag set with only certain morphological features. Therefore, researchers have proposed tag sets for MSA whose size is similar to that of the English PTB tag set, as this has proven to be a useful size computationally. Citation Sentence: A more detailed discussion of the various available Arabic tag sets can be found in Habash ( 2010 ) . Context after the citation: The following are the various tag sets we use in this article: (a) the core POS tag sets CORE44 and the newly introduced CORE12; (b) CATiB Treebank tag set (CATIB6) (Habash and Roth 2009) and its newly introduced extension of CATIBEX created using simple regular expressions on word form, indicating particular morphemes such as the prefix JI Al+ or the suffix v' +wn; this tag set is the best-performing tag set for Arabic on predicted values as reported in Section 4; (c) the PATB full tag set with complete morphological tag (BW) (Buckwalter 2004); and two extensions of the PATB reduced tag set (PENN POS, a.k.a. RTS, size 24 [Diab, Hacioglu, and Jurafsky 2004]), both outperforming it: (d) Kulick, Gabbard, and Marcus (2006)âs tag set (KULICK), size 43, one of whose most important extensions is the marking of the definite article clitic, and (e) Diab and Benajibaâs (in preparation) EXTENDED RTS tag set (ERTS), which marks gender, number, and definiteness, size 134.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1009
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: But while Bod's estimator obtains state-of-the-art results on the WSJ, comparable to Charniak (2000) and Collins (2000), Bonnema et al.'s estimator performs worse and is comparable to Collins (1996). We show that these PCFG-reductions result in a 60 times speedup in processing time w.r.t. Bod (2001, 2003). This paper presents the first published results with Goodman's PCFG-reductions of both Bonnema et al.'s (1999) and Bod's (2001) estimators on the WSJ. Citation Sentence: But while Bod 's estimator obtains state-of-the-art results on the WSJ , comparable to Charniak ( 2000 ) and Collins ( 2000 ) , Bonnema et al. 's estimator performs worse and is comparable to Collins ( 1996 ) . Context after the citation: In the second part of this paper, we extend our experiments with a new notion of the best parse tree. Most previous notions of best parse tree in DOP1 were based on a probabilistic metric, with Bod (2000b) as a notable exception, who used a simplicity metric based on the shortest derivation. We show that a combination of a probabilistic and a simplicity metric, which chooses the simplest parse from the n likeliest parses, outperforms the use of these metrics alone.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:101
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The high Dirichlet priors are chosen to prevent sparsity in topic distributions, while the other parameters are selected as the best from Hoffman et al. (2010). We do not optimize these hyperparameters or vary them over time. In all settings, we fix all Dirichlet priors at 0.1, use a learning rate 0.7, and use minibatch sizes of 1024 documents. Citation Sentence: The high Dirichlet priors are chosen to prevent sparsity in topic distributions , while the other parameters are selected as the best from Hoffman et al. ( 2010 ) . Context after the citation: In order to optimize the number of topics K, we run five trials of each modality for 2000 iterations for K = 150, 100, 150, 200, 2501 (a total of 25 runs per setup). We select the value or K for each model which minimizes the average perplexity estimate over the five trials.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1010
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The resulting speech understanding system is called the Voice Natural Language Computer with Expectation (VNLCE, Fink 1983). The expectation system provided the intermediate processing between the errorful output of the speech recognizer and the deep semantics of NLC. An off-the-shelf speech recognition device, a Nippon Electric Corporation DP-200, was added to an existing natural language processing system, the Natural Language Computer (NLC) (Ballard 1979, Biermann and Ballard 1980). Citation Sentence: The resulting speech understanding system is called the Voice Natural Language Computer with Expectation ( VNLCE , Fink 1983 ) . Context after the citation: [The current system should be distinguished from an earlier voice system (VNLC, Biermann et al. 1985), which had no expectation and which handled discrete speech where a 300 millisecond pause must follow each word.] It should be emphasized, of course, that the central issue here is the study of expectation mechanisms and the details of the design decisions could have been made in rather different ways. Thus one could have implemented expectation error correction with a typed input system or with a speech input system that integrates voice signal processing with higher level functions in a way not possible with a commercial recognizer. This implementation shows only one way in which the functions P, Predicts, Mergeable, and Merge can be constructed to achieve expectation capabilities.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1011
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: In our previous papers (Zhang and Clark 2011; Zhang, Blackwood, and Clark 2012), we applied a set of beams to this structure, which makes it similar to the data structure used for phrase-based MT decoding (Koehn 2010). Second, although we use the term chart, the structure for accepted hypotheses is not a dynamic programming chart in the same way as for the parsing problem. First, the parsing problem has a fixed word order and is considerably simpler than the word ordering problem we are tackling. Citation Sentence: In our previous papers ( Zhang and Clark 2011 ; Zhang , Blackwood , and Clark 2012 ) , we applied a set of beams to this structure , which makes it similar to the data structure used for phrase-based MT decoding ( Koehn 2010 ) . Context after the citation: However, we will show later that this structure is unnecessary when the model has more discriminative power, and a conceptually simpler single beam can be used. We will also investigate the possibility of applying dynamic-programming-style pruning to the chart. We now give an overview of the training algorithm, which is crucial to both the speed and accuracy of the resulting decoder. CCGBank (Hockenmaier and Steedman 2007) is used to train the model.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1012
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: A number of speech understanding systems have been developed during the past fifteen years (Barnett et al. 1980, Dixon and Martin 1979, Erman et al. 1980, Haton and Pierrel 1976, Lea 1980, Lowerre and Reddy 1980, Medress 1980, Reddy 1976, Walker 1978, and Wolf and Woods 1980). Citation Sentence: A number of speech understanding systems have been developed during the past fifteen years ( Barnett et al. 1980 , Dixon and Martin 1979 , Erman et al. 1980 , Haton and Pierrel 1976 , Lea 1980 , Lowerre and Reddy 1980 , Medress 1980 , Reddy 1976 , Walker 1978 , and Wolf and Woods 1980 ) . Context after the citation: Most of these efforts concentrated on the interaction between low level information sources from a speech recognizer and a natural language processor to discover the meaning of an input sentence. While some of these systems did exhibit expectation capabilities at the sentence level, none acquired dialogues of the kind described here for the sake of dialogue level expectation and error correction. A detailed description of the kinds of expectation mechanisms appearing in these systems appears in Fink (1983). The problem of handling ill-formed input has been studied by Carbonell and Hayes (1983), Granger (1983), Jensen et al. (1983), Kwasny and Sondheimer (1981), Riesbeck and Schank (1976), Thompson (1980), Weischedel and Black (1980), and Weischedel and Sondheimer (1983).	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1013
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Another line of research approaches grounded language knowledge by augmenting distributional approaches of word meaning with perceptual information (Andrews et al., 2009; Steyvers, 2010; Feng and Lapata, 2010b; Bruni et al., 2011; Silberer and Lapata, 2012; Johns and Jones, 2012; Bruni et al., 2012a; Bruni et al., 2012b; Silberer et al., 2013). Some efforts have tackled tasks such as automatic image caption generation (Feng and Lapata, 2010a; Ordonez et al., 2011), text illustration (Joshi et al., 2006), or automatic location identification of Twitter users (Eisenstein et al., 2010; Wing and Baldridge, 2011; Roller et al., 2012). Others provide automatic mappings of natural language instructions to executable actions, such as interpreting navigation directions (Chen and Mooney, 2011) or robot commands (Tellex et al., 2011; Matuszek et al., 2012). Citation Sentence: Another line of research approaches grounded language knowledge by augmenting distributional approaches of word meaning with perceptual information ( Andrews et al. , 2009 ; Steyvers , 2010 ; Feng and Lapata , 2010b ; Bruni et al. , 2011 ; Silberer and Lapata , 2012 ; Johns and Jones , 2012 ; Bruni et al. , 2012a ; Bruni et al. , 2012b ; Silberer et al. , 2013 ) . Context after the citation: Although these approaches have differed in model definition, the general goal in this line of research has been to enhance word meaning with perceptual information in order to address one of the most common criticisms of distributional semantics: that the âmeaning of words is entirely given by other wordsâ (Bruni et al., 2012b). In this paper, we explore various ways to integrate new perceptual information through novel computational modeling of this grounded knowledge into a multimodal distributional model of word meaning. The model we rely on was originally developed by Andrews et al. (2009) and is based on a generalization of Latent Dirichlet Allocation.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1014
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: In particular, since we treat each individual speech within a debate as a single âdocumentâ, we are considering a version of document-level sentiment-polarity classification, namely, automatically distinguishing between positive and negative documents (Das and Chen, 2001; Pang et al., 2002; Turney, 2002; Dave et al., 2003). Task properties Determining whether or not a speaker supports a proposal falls within the realm of sentiment analysis, an extremely active research area devoted to the computational treatment of subjective or opinion-oriented language (early work includes Wiebe and Rapaport (1988), Hearst (1992), Sack (1994), and Wiebe (1994); see Esuli (2006) for an active bibliography). Note that from an experimental point of view, this is a very convenient problem to work with because we can automatically determine ground truth (and thus avoid the need for manual annotation) simply by consulting publicly available voting records. Citation Sentence: In particular , since we treat each individual speech within a debate as a single `` document '' , we are considering a version of document-level sentiment-polarity classification , namely , automatically distinguishing between positive and negative documents ( Das and Chen , 2001 ; Pang et al. , 2002 ; Turney , 2002 ; Dave et al. , 2003 ) . Context after the citation: Most sentiment-polarity classifiers proposed in the recent literature categorize each document independently. A few others incorporate various measures of inter-document similarity between the texts to be labeled (Agarwal and Bhattacharyya, 2005; Pang and Lee, 2005; Goldberg and Zhu, 2006). Many interesting opinion-oriented documents, however, can be linked through certain relationships that occur in the context of evaluative discussions. For example, we may find textual4 evidence of a high likelihood of agreement be4Because we are most interested in techniques applicable across domains, we restrict consideration to NLP aspects of the problem, ignoring external problem-specific information.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1015
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: A number of alignment techniques have been proposed, varying from statistical methods (Brown et al., 1991; Gale and Church, 1991) to lexical methods (Kay and Roscheisen, 1993; Chen, 1993). Aligning English-Chinese parallel texts is already very difficult because of the great differences in the syntactic structures and writing systems of the two languages. Some are highly parallel and easy to align while others can be very noisy. Citation Sentence: A number of alignment techniques have been proposed , varying from statistical methods ( Brown et al. , 1991 ; Gale and Church , 1991 ) to lexical methods ( Kay and Roscheisen , 1993 ; Chen , 1993 ) . Context after the citation: The method we adopted is that of Simard et al. (1992). Because it considers both length similarity and cognateness as alignment criteria, the method is more robust and better able to deal with noise than pure length-based methods. Cognates are identical sequences of characters in corresponding words in two languages. They are commonly found in English and French.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1016
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: (Michiels (1982) contains further description and discussion of LDOCE.) Most prominent among these are the rich grammatical subcategorisations of the 60,000 entries, the large amount of information concerning phrasal verbs, noun compounds and idioms, the individual subject, collocational and semantic codes for the entries and the consistent use of a controlled 'core' vocabulary in defining the words throughout the dictionary. We chose to employ LDOCE as the machine readable source to aid the development of a substantial lexicon because this dictionary has several properties which make it uniquely appropriate for use as the core knowledge base of a natural language processing system. Citation Sentence: ( Michiels ( 1982 ) contains further description and discussion of LDOCE . ) Context after the citation: In this paper we focus on the exploitation of the LDOCE grammar coding system; Alshawi et al. (1985) and Alshawi (1987) describe further research in Cambridge utilising different types of information available in LDOCE. The information available in the dictionary is both very rich and diverse, but also typically only semiformalised, as it is intended for human, rather than machine, interpetation. As a consequence the programs we are developing, both to restructure and to exploit this information, need to undergo constant revision as they are being used. The system we describe is not intended for off-line use, where one might attempt to derive, completely automatically, a lexicon for natural language analysis.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1017
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Juola (1994, 1997) assumes that words ending in -ed are verbs. That is, using the marker hypothesis method of segmentation, smaller aligned segments can be extracted from the phrasal lexicon without recourse to any detailed parsing techniques or complex co-ocurrence measures. <QUANT> all : tous <PREP> of : dâ <LEX> uses : usages <LEX> asbestos : asbeste Citation Sentence: Juola ( 1994 , 1997 ) assumes that words ending in - ed are verbs . Context after the citation: However, given that verbs are not a closed class, in our approach we do not mark chunks beginning with a verb with any marker category. Instead, we take advantage of the fact that the initial phrasal chunks correspond to rule right-hand sides. That is, for a rule in the Penn Treebank VP â> VBG, NP, PP, we are certain (if the annotators have done their job correctly) that the first word in each of the strings corresponding to this right-hand side is a VBG, that is, a present participle. Given this information, in such cases we tag such words with the <LEX> tag.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1018
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: For the development of these lists we used a collection of texts of about 300,000 words derived from the New York Times (NYT) corpus that was supplied as training data for the 7th Message Understanding Conference (MUC-7) (Chinchor 1998). These four lists can be acquired completely automatically from raw (unlabeled) texts. â¢ frequent proper name â¢ abbreviation (as opposed to regular word) Citation Sentence: For the development of these lists we used a collection of texts of about 300,000 words derived from the New York Times ( NYT ) corpus that was supplied as training data for the 7th Message Understanding Conference ( MUC-7 ) ( Chinchor 1998 ) . Context after the citation: We used these texts because the approach described in this article was initially designed to be part of a named-entity recognition system (Mikheev, Grover, and Moens 1998) developed for MUC-7. Although the corpus size of 300,000 words can be seen as large, the fact that this corpus does not have to be annotated in any way and that a corpus of similar size can be easily collected from on-line sources (including the Internet) makes this resource cheap to obtain. The first list on which our method relies is a list of common words. This list includes common words for a given language, but no supplementary information such as POS or morphological information is required to be present in this list.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1019
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: conclusion Context before the citation: In multi-party discussion people usually mention each otherâs name for the purpose of disentanglement (Elsner and Charniak, 2008). Another possibly critical feature is the âmention of namesâ. The best baseline âSpeakerâ in Table 4 also favours this claim. Citation Sentence: In multi-party discussion people usually mention each other 's name for the purpose of disentanglement ( Elsner and Charniak , 2008 ) . Context after the citation: In our corpus we found 175 instances where a participant mentions other participantâs name. In addition to these, âSubject of the emailâ, âtopic-shift cue wordsâ can also be beneficial for a model. As a next step for this research, we will investigate how to exploit these features in our methods. We are also interested in the near future to transfer our approach to other similar domains by hierarchical Bayesian multi-task learning and other domain adaptation methods.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:102
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: We would also provide a user facility for choosing the right licence for every lexia, following the model of Creative Commons licences (Lessig, 2004). We want to model views as dynamic objects the creation of context will be still arbitrary, but changes are very easily. With our typology of links, we aim to solve the framing problem as defined in Section 1.2. Citation Sentence: We would also provide a user facility for choosing the right licence for every lexia , following the model of Creative Commons licences ( Lessig , 2004 ) . Context after the citation:	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1020
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: The language grounding problem has received significant attention in recent years, owed in part to the wide availability of data sets (e.g. Flickr, Von Ahn (2006)), computing power, improved computer vision models (Oliva and Torralba, 2001; Lowe, 2004; Farhadi et al., 2009; Parikh and Grauman, 2011) and neurological evidence of ties between the language, perceptual and motor systems in the brain (PulvermÂ¨uller et al., 2005; Tettamanti et al., 2005; Aziz-Zadeh et al., 2006). Citation Sentence: The language grounding problem has received significant attention in recent years , owed in part to the wide availability of data sets ( e.g. Flickr , Von Ahn ( 2006 ) ) , computing power , improved computer vision models ( Oliva and Torralba , 2001 ; Lowe , 2004 ; Farhadi et al. , 2009 ; Parikh and Grauman , 2011 ) and neurological evidence of ties between the language , perceptual and motor systems in the brain ( Pulverm Â¨ uller et al. , 2005 ; Tettamanti et al. , 2005 ; Aziz-Zadeh et al. , 2006 ) . Context after the citation: Many approaches to multimodal research have succeeded by abstracting away raw perceptual information and using high-level representations instead. Some works abstract perception via the usage of symbolic logic representations (Chen et al., 2010; Chen and Mooney, 2011; Matuszek et al., 2012; Artzi and Zettlemoyer, 2013), while others choose to employ concepts elicited from psycholinguistic and cognition studies. Within the latter category, the two most common representations have been association norms, where subjects are given a 1http://stephenroller.com/research/ emnlp13	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1021
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: conclusion Context before the citation: There has been some controversy, at least for simple stemmers (Lovins, 1968; Porter, 1980), about the effectiveness of morphological analysis for document retrieval (Harman, 1991; Krovetz, 1993; Hull, 1996). Citation Sentence: There has been some controversy , at least for simple stemmers ( Lovins , 1968 ; Porter , 1980 ) , about the effectiveness of morphological analysis for document retrieval ( Harman , 1991 ; Krovetz , 1993 ; Hull , 1996 ) . Context after the citation: The key for quality improvement seems to be rooted mainly in the presence or absence of some form of dictionary. Empirical evidence has been brought forward that inflectional and/or derivational stemmers augmented by dictionaries indeed perform substantially better than those without access to such lexical repositories (Krovetz, 1993; Kraaij and Pohlmann, 1996; Tzoukermann et al., 1997). This result is particularly valid for natural languages with a rich morphology â both in terms of derivation and (single-word) composition. Document retrieval in these languages suffers from serious performance degradation with the stemmingonly query-term-to-text-word matching paradigm.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1022
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: We have yet to import such a constraint into our model, but we plan to do so in the near future using the weighted majority algorithm (Littlestone and Warmuth 1992). That is, we want to allocate a larger portion of the probability space to the phrasal and marker lexicons than to the generalized or wordlevel lexicons. We would like to incorporate into our model a procedure whereby translation chunks extracted from the phrasal and marker lexicons are more highly regarded than those constructed by inserting words from the word-level lexicon into generalized marker chunks. Citation Sentence: We have yet to import such a constraint into our model , but we plan to do so in the near future using the weighted majority algorithm ( Littlestone and Warmuth 1992 ) . Context after the citation:	FutureWork	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1023
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Generally speaking, we find that the personal public diary metaphor behind blogs (McNeill, 2005) may bring to an unsatisfactory representation of the context. So we have tested and compared the most used tools available for blogging: Bloggers, WordPress, MovableType and LiveJournal. While wikis have spread from a detailed design (Cunningham and Leuf, 2001), unfortunately blogs have not been designed under a model. Citation Sentence: Generally speaking , we find that the personal public diary metaphor behind blogs ( McNeill , 2005 ) may bring to an unsatisfactory representation of the context . Context after the citation: The only way to retrieve information is through a search engine or a calendar, i.e. the date of the âpostâ â a lexia in the jargon of bloggers. Moreover, we use some new web applications to take and share notes or to browser everyoneâs bookmarks, e.g. del.icio.us. Mostly, these web applications oriented to writing give a strong emphasis on collaboration and sharing.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1024
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Most probabilistic translation model reestimation algorithms published to date are variations on the theme proposed by Brown et al. (1993b). Citation Sentence: Most probabilistic translation model reestimation algorithms published to date are variations on the theme proposed by Brown et al. ( 1993b ) . Context after the citation: These models involve conditional probabilities, but they can be compared to symmetric models if the latter are normalized by the appropriate marginal distribution. I shall review these models using the notation in Table 1. 4.3.1 Models Using Only Co-occurrence Information. Brown and his colleagues em-	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1025
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: A number of proposals in the 1990s deliberately limited the extent to which they relied on domain and/or linguistic knowledge and reported promising results in knowledge-poor operational environments (Dagan and Itai 1990, 1991; Lappin and Leass 1994; Nasukawa 1994; Kennedy and Boguraev 1996; Williams, Harvey, and Preston 1996; Baldwin 1997; Mitkov 1996, 1998b). However, the pressing need for the development of robust and inexpensive solutions to meet the demands of practical NLP systems encouraged many researchers to move away from extensive domain and linguistic knowledge and to embark instead upon knowledge-poor anaphora resolution strategies. Much of the earlier work in anaphora resolution heavily exploited domain and linguistic knowledge (Sidner 1979; Carter 1987; Rich and LuperFoy 1988; Carbonell and Brown 1988), which was difficult both to represent and to process, and which required considerable human input. Citation Sentence: A number of proposals in the 1990s deliberately limited the extent to which they relied on domain and/or linguistic knowledge and reported promising results in knowledge-poor operational environments ( Dagan and Itai 1990 , 1991 ; Lappin and Leass 1994 ; Nasukawa 1994 ; Kennedy and Boguraev 1996 ; Williams , Harvey , and Preston 1996 ; Baldwin 1997 ; Mitkov 1996 , 1998b ) . Context after the citation: The drive toward knowledge-poor and robust approaches was further motivated by the emergence of cheaper and more reliable corpus-based NLP tools such as partof-speech taggers and shallow parsers, alongside the increasing availability of corpora and other NLP resources (e.g., ontologies). In fact, the availability of corpora, both raw and annotated with coreferential links, provided a strong impetus to anaphora resolu- tion with regard to both training and evaluation. Corpora (especially when annotated) are an invaluable source not only for empirical research but also for automated learning (e.g., machine learning) methods aiming to develop new rules and approaches; they also provide an important resource for evaluation of the implemented approaches.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1026
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The significance testing is performed by paired bootstrap re-sampling (Koehn, 2004). In our experiments the translation performances are measured by case-insensitive BLEU4 metric (Papineni et al., 2002) and we use mtevalv13a.pl as the evaluation tool. We train a 4-gram language model on the Xinhua portion of the English Gigaword corpus using the SRILM Toolkits (Stolcke, 2002) with modified Kneser-Ney smoothing (Chen and Goodman, 1998). Citation Sentence: The significance testing is performed by paired bootstrap re-sampling ( Koehn , 2004 ) . Context after the citation: We use an in-house developed hierarchical phrase-based translation (Chiang, 2005) as our baseline system, and we denote it as In-Hiero. To obtain satisfactory baseline performance, we tune InHiero system for 5 times using MERT, and then se- (MERT). NIST05 is the set used to tune A for MBUU and EBUU, and NIST06 and NIST08 are test sets.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1027
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The system is implemented based on (Galley et al., 2006) and (Marcu et al. 2006). The translation system used for testing the effectiveness of our U-trees is our in-house stringto-tree system (abbreviated as s2t). To create the baseline system, we use the opensource Joshua 4.0 system (Ganitkevitch et al., 2012) to build a hierarchical phrase-based (HPB) system, and a syntax-augmented MT (SAMT) 11 system (Zollmann and Venugopal, 2006) respectively. Citation Sentence: The system is implemented based on ( Galley et al. , 2006 ) and ( Marcu et al. 2006 ) . Context after the citation: In the system, we extract both the minimal GHKM rules (Galley et al., 2004), and the rules of SPMT Model 1 (Galley et al., 2006) with phrases up to length L=5 on the source side. We then obtain the composed rules by composing two or three adjacent minimal rules. To build the above s2t system, we first use the parse tree, which is generated by parsing the English side of the bilingual data with the Berkeley parser (Petrov et al., 2006). Then, we binarize the English parse trees using the head binarization approach (Wang et al., 2007) and use the resulting binary parse trees to build another s2t system.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1028
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: IGEN uses standard chart generation techniques (Kay, 1996) in its base generator to efficiently produce generation candidates. The ranks are determined by the similarity to the closest instances and the highest ranked sentence is chosen as the final generation output. The ranker scores these candidates according to a similarity metric which measures their distance to the elements in the instance base. Citation Sentence: IGEN uses standard chart generation techniques ( Kay , 1996 ) in its base generator to efficiently produce generation candidates . Context after the citation: The set of candidates forms a subset of the chart edges in that they comprise all edges of syntactic category S. The particular advantage of a bottom-up generation algorithm is that it allows our surface-based ranker to score edges as soon as they are built.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1029
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: Fortunately, indirect associations are usually not difficult to identify, because they tend to be weaker than the direct associations on which they are based (Melamed, 1996c). Models of translational equivalence that are ignorant of indirect associations have "a tendency ... to be confused by collocates" (Dagan et al., 1993). The arrow connecting vk and ukÂ±i in Figure 1 represents an indirect association, since the association between vk and ukÂ±i arises only by virtue of the association between each of them and uk . Citation Sentence: Fortunately , indirect associations are usually not difficult to identify , because they tend to be weaker than the direct associations on which they are based ( Melamed , 1996c ) . Context after the citation: The majority of indirect associations can be filtered out by a simple competition heuristic: Whenever several word tokens ui in one half of the bitext co-occur with a particular word token v in the other half of the bitext, the word that is most likely to be v's translation is the one for which the likelihood L(u, v) of translational equivalence is highest. The competitive linking algorithm implements this heuristic: 1. Discard all likelihood scores for word types deemed unlikely to be mutual translations, i.e. all L(u, v) < 1.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:103
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Position, subcat frame, phrase type, first word, last word, subcat frame+, predicate, path, head word and its POS, predicate + head word, predicate + phrase type, path to BA and BEI, verb class 3 , verb class + head word, verb class + phrase type, from Xue (2008). Head word POS, Head Word of Prepositional Phrases, Constituent tree distance, from Pradhan etc. (2004). The candidate feature templates include: Voice from Sun and Jurafsky (2004). Citation Sentence: Position , subcat frame , phrase type , first word , last word , subcat frame + , predicate , path , head word and its POS , predicate + head word , predicate + phrase type , path to BA and BEI , verb class 3 , verb class + head word , verb class + phrase type , from Xue ( 2008 ) . Context after the citation: 3 It is a bit different from Xue (2008), since we didnât use the syntactic alternation information. predicate POS, first word + last word, phrase type of the sibling to the left, phrase type of the sibling to the right, verb + subcate frame+, verb POS + subcat frame+, the amount of VPs in path, phrase type + phrase type of parent node, which can be easily understood by name. voice position, indicates whether the voice marker (BA, BEI) is before or after the constituent in focus. subcat frame*, the rule that expands the parent node of the constituent in focus.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1030
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: The psycholinguistic studies of Martin (1970), Allen (1975), Hillinger et al. (1976), Grosjean et al. (1979), Dommergues and Grosjean (1983), and Gee and Grosjean (1983), responding to the idea of readjusted syntax as the source of prosodic phrasing, show that grammatical structure, even if readjusted, is not in itself a reliable predictor of prosodic phrasing: mismatches between syntax and prosody occur often and systematically, and can be related to specific nonsyntactic factors such as length and word frequency. Citation Sentence: The psycholinguistic studies of Martin ( 1970 ) , Allen ( 1975 ) , Hillinger et al. ( 1976 ) , Grosjean et al. ( 1979 ) , Dommergues and Grosjean ( 1983 ) , and Gee and Grosjean ( 1983 ) , responding to the idea of readjusted syntax as the source of prosodic phrasing , show that grammatical structure , even if readjusted , is not in itself a reliable predictor of prosodic phrasing : mismatches between syntax and prosody occur often and systematically , and can be related to specific nonsyntactic factors such as length and word frequency . Context after the citation: For example, although prosodic boundaries between subject and verb do occur, there also exist prosodic patterns in which the boundary comes between the verb and object, i.e., the data reveal both X(VY) and (XV)Y groupings. Grosjean et al. (1979) claims that such mismatches are due for the most part to constituent length, which interacts with grammatical structure and, in some cases, overrides it. Thus syntactic and prosodic structure match when the major constituents of a sentence are roughly equal in length; for example, the main prosodic phrase break corresponds to the subject-predicate boundary in Waiters who remember well II serve orders correctly. Discrepancies in length throw constituents off balance, and so prosodic phrasing will cross constituent boundaries in order to give the phrases similar lengths; this is the case in Chickens were eating II the remaining green vegetables, where the subject-predicate boundary finds no prosodic correspondent.4 The most explicit version of this approach is the analysis presented in Gee and Grosjean (1983) (henceforth G&G).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1031
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: A number of studies (e.g., Hildebrandt, Katz, and Lin 2004) have pointed out shortcomings of the original nugget scoring model, although a number of these issues have been recently addressed (Lin and Demner-Fushman 2005a, 2006b). Nugget F-score has been employed as a metric in the TREC question-answering track since 2003, to evaluate so-called definition and âotherâ questions (Voorhees 2003). Recently, there is a growing consensus that an evaluation methodology based on the notion of âinformation nuggetsâ may provide an appropriate framework for assessing the quality of answers to complex questions. Citation Sentence: A number of studies ( e.g. , Hildebrandt , Katz , and Lin 2004 ) have pointed out shortcomings of the original nugget scoring model , although a number of these issues have been recently addressed ( Lin and Demner-Fushman 2005a , 2006b ) . Context after the citation: However, adaptation of the nugget evaluation methodology to a domain as specific as clinical medicine is an endeavor that has yet to be undertaken.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1032
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Nasr (1997) reported that the translation lexicon that our model induced from this tiny bitext accounted for 30% of the word types with precision between 84% and 90%. The model was also used to induce a translation lexicon from a 6200-word corpus of French/English weather reports. translation lexicons with 94% precision and 30% recall, when trained on French/English software manuals totaling about 400,000 words. Citation Sentence: Nasr ( 1997 ) reported that the translation lexicon that our model induced from this tiny bitext accounted for 30 % of the word types with precision between 84 % and 90 % . Context after the citation: Recall drops when there is less training data, because the model refuses to make predictions that it cannot make with confidence. For many applications, this is the desired behavior.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1033
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Kinyon and Prolo (2002) describe a simple tool which uses fine-grained rules to identify the arguments of verb occurrences in the Penn-II Treebank. Approaches using treebank-based data as a source for subcategorization information, such as ours, do not predefine the frames to be extracted but rather learn them from the data. The extensive evaluation carried out by Schulte im Walde will be discussed in greater detail in Section 6. Citation Sentence: Kinyon and Prolo ( 2002 ) describe a simple tool which uses fine-grained rules to identify the arguments of verb occurrences in the Penn-II Treebank . Context after the citation: This is made possible by manual examination of more than 150 different sequences of syntactic and functional tags in the treebank. Each of these sequences was categorized as a modifier or argument. Arguments were then mapped to traditional syntactic functions. For example, the tag sequence NP-SBJ denotes a mandatory argument, and its syntactic function is subject.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1034
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The contextual interpreter then uses a reference resolution approach similar to Byron (2002), and an ontology mapping mechanism (Dzikovska et al., 2008a) to produce a domain-specific semantic representation of the studentâs output. The parser provides a domain-independent semantic representation including high-level word senses and semantic role labels. We use the TRIPS dialogue parser (Allen et al., 2007) to parse the utterances. Citation Sentence: The contextual interpreter then uses a reference resolution approach similar to Byron ( 2002 ) , and an ontology mapping mechanism ( Dzikovska et al. , 2008a ) to produce a domain-specific semantic representation of the student 's output . Context after the citation: Utterance content is represented as a set of extracted objects and relations between them. Negation is supported, together with a heuristic scoping algorithm. The interpreter also performs basic ellipsis resolution. For example, it can determine that in the answer to the question âWhich bulbs will be on and which bulbs will be off in this diagram?â	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1035
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Zhu (2005) maintains a survey of this area. Notable early papers on graph-based semisupervised learning include Blum and Chawla (2001), Bansal et al. (2002), Kondor and Lafferty (2002), and Joachims (2003). inter-document references in the form of hyperlinks (Agrawal et al., 2003). Citation Sentence: Zhu ( 2005 ) maintains a survey of this area . Context after the citation: Recently, several alternative, often quite sophisticated approaches to collective classification have been proposed (Neville and Jensen, 2000; Lafferty et al., 2001; Getoor et al., 2002; Taskar et al., 2002; Taskar et al., 2003; Taskar et al., 2004; McCallum and Wellner, 2004). It would be interesting to investigate the application of such methods to our problem. However, we also believe that our approach has important advantages, including conceptual simplicity and the fact that it is based on an underlying optimization problem that is provably and in practice easy to solve.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1036
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: The principle of maximum entropy states that when one searches among probability distributions that model the observed data (evidence), the preferred one is the one that maximizes the entropy (a measure of the uncertainty of the model) (Berger et al., 1996). We use a maximum entropy Markov model (MEMM) classifier. We assign to each token in the text a label indicating whether it starts a specific mention, is inside a specific mention, or is outside any mentions. Citation Sentence: The principle of maximum entropy states that when one searches among probability distributions that model the observed data ( evidence ) , the preferred one is the one that maximizes the entropy ( a measure of the uncertainty of the model ) ( Berger et al. , 1996 ) . Context after the citation: One big advantage of this approach is that it can combine arbitrary and diverse types of information in making a classification decision. Our mention detection system predicts the four labels types associated with a mention through a cascade approach. It first predicts the boundary and the main entity type for each mention. Then, it uses the information regarding the type and boundary in different second-stage classifiers to predict the subtype, the mention level, and the mention class.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1037
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: In the context of word alignment, Deng and Byrne (2005) use a state-duration HMM in order to model word-to-phrase translations. The idea of introducing constraints over a model to better guide the learning process has appeared before. Citation Sentence: In the context of word alignment , Deng and Byrne ( 2005 ) use a state-duration HMM in order to model word-to-phrase translations . Context after the citation: The fertility of each source word is implicitly encoded in the durations of the HMM states. Without any restrictions, likelihood prefers to always use longer phrases and the authors try to control this behavior by multiplying every transition probability by a constant ri > 1. This encourages more transitions and hence shorter phrases. For the task of unsupervised dependency parsing, Smith and Eisner (2006) add a constraint of the form âthe average length of dependencies should be Xâ to capture the locality of syntax (at least half of the dependencies are between adjacent words), using a scheme they call structural annealing.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1038
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: With respect to the focus on function words, our reordering model is closely related to the UALIGN system (Hermjakob, 2009). These related models treat their lexical items as latent variables to be estimated from training data, while our model uses a fixed set of lexical items that correspond to the class of function words. Our reordering model is closely related to the model proposed by Zhang and Gildea (2005; 2006; 2007a), with respect to conditioning the reordering predictions on lexical items. Citation Sentence: With respect to the focus on function words , our reordering model is closely related to the UALIGN system ( Hermjakob , 2009 ) . Context after the citation: However, UALIGN uses deep syntactic analysis and hand-crafted heuristics in its model.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1039
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: The work of Sarkar (2001) and Steedman, Osborne, et al. (2003) suggests that co-training can be helpful for statistical parsing. Another technique for making better use of unlabeled data is cotraining (Blum and Mitchell 1998), in which two sufficiently different learners help each other learn by labeling training data for one another. This approach assumes that there are enough existing labeled data to train the individual parsers. Citation Sentence: The work of Sarkar ( 2001 ) and Steedman , Osborne , et al. ( 2003 ) suggests that co-training can be helpful for statistical parsing . Context after the citation: Pierce and Cardie (2001) have shown, in the context of base noun identification, that combining sample selection and cotraining can be an effective learning framework for large-scale training. Similar approaches are being explored for parsing (Steedman, Hwa, et al. 2003; Hwa et al. 2003).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:104
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: conclusion Context before the citation: We could also introduce new variables, e.g., nonterminal refinements (Matsuzaki et al., 2005), or secondary links Mid (not constrained by TREE/PTREE) that augment the parse with representations of control, binding, etc. (Sleator and Temperley, 1993; Buch-Kromann, 2006). For projective parsing, it is significantly faster than exact dynamic programming, at the cost of small amounts of search error, We are interested in extending these ideas to phrase-structure and lattice parsing, and in trying other higher-order features, such as those used in parse reranking (Charniak and Johnson, 2005; Huang, 2008) and history-based parsing (Nivre and McDonald, 2008). Belief propagation improves non-projective dependency parsing with features that would make exact inference intractable. Citation Sentence: We could also introduce new variables , e.g. , nonterminal refinements ( Matsuzaki et al. , 2005 ) , or secondary links Mid ( not constrained by TREE/PTREE ) that augment the parse with representations of control , binding , etc. ( Sleator and Temperley , 1993 ; Buch-Kromann , 2006 ) . Context after the citation: Other parsing-like problems that could be attacked with BP appear in syntax-based machine translation. Decoding is very expensive with a synchronous grammar composed with an n-gram language model (Chiang, 2007)âbut our footnote 10 suggests that BP might incorporate a language model rapidly. String alignment with synchronous grammars is quite expensive even for simple synchronous formalisms like ITG (Wu, 1997)âbut Duchi et al. (2007) show how to incorporate bipartite matching into max-product BP. Finally, we can take advantage of improvements to BP proposed in the context of other applications.	FutureWork	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1040
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Selectional Preferences have also been a recent focus of researchers investigating the learning of paraphrases and inference rules (Pantel et al., 2007; Roberto et al., 2007). Similarity-smoothed models are simple to compute, potentially adaptable to new domains, and require no manually-compiled resources such as WordNet. Erk (2007) compared a number of techniques for creating similar-word sets and found that both the Jaccard coefficient and Lin (1998a)âs information-theoretic metric work best. Citation Sentence: Selectional Preferences have also been a recent focus of researchers investigating the learning of paraphrases and inference rules ( Pantel et al. , 2007 ; Roberto et al. , 2007 ) . Context after the citation: Inferences such as â[X wins Y] â [X plays Y]â are only valid for certain arguments X and Y. We follow Pantel et al. (2007) in using automatically-extracted semantic classes to help characterize plausible arguments. Discriminative techniques are widely used in NLP and have been applied to the related tasks of word prediction and language modeling. Even-Zohar and Roth (2000) use a classifier to predict the most likely word to fill a position in a sentence (in their experiments: a verb) from a set of candidates (sets of verbs), by inspecting the context of the target token (e.g., the presence or absence of a particular nearby word in the sentence).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1041
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: Previously LDA has been successfully used to infer unsupervised joint topic distributions over words and feature norms together (Andrews et al., 2009; Silberer and Lapata, 2012). Our experiments are based on the multimodal extension of Latent Dirichlet Allocation developed by Andrews et al. (2009). Citation Sentence: Previously LDA has been successfully used to infer unsupervised joint topic distributions over words and feature norms together ( Andrews et al. , 2009 ; Silberer and Lapata , 2012 ) . Context after the citation: It has also been shown to be useful in joint inference of text with visual attributes obtained using visual classifiers (Silberer et al., 2013). These multimodal LDA models (hereafter, mLDA) have been shown to be qualitatively sensible and highly predictive of several psycholinguistic tasks (Andrews et al., 2009). However, prior work using mLDA is limited to two modalities at a time. In this section, we describe bimodal mLDA and define two methods for extending it to three or more modalities.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1042
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Something like this approach is in fact used in some systems (e.g., Elhadad and Robin 1992; PenMan 1989; Hovy 1988a). The planner could supply whatever information is needed to drive the network.' If that were the only goal, we could dispense with the feedback mechanism and simply design some sort of discrimination network (or similar device) to test various features of the information being expressed. Citation Sentence: Something like this approach is in fact used in some systems ( e.g. , Elhadad and Robin 1992 ; PenMan 1989 ; Hovy 1988a ) . Context after the citation: The problem with the discrimination network approach is that it can't handle the range of examples shown here without violating the modularity of the generator. The examples shown here have involved linguistic choices that depend on the plan structure, on other actions in the plan, and on preconditions of the plan. Furthermore, the plan dependencies can involve a nontrivial amount of reasoning (e.g., the example immediately above, which involves the connection between sunshine and enjoying 33 Of course, that still wouldn't allow for examples like the one in Section 4.2, where the planner modifies the plan in response to the work of the linguistic component. being outdoors).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1043
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: The emphasis on narrativity takes into account the use of blogs as public diaries on the web, that is still the main current interpretation of this literary genre, or metagenre (McNeill, 2005). We believe that this name is clearly understable to every people educated in a European-based culture, and this is why we have chosen it. It resembles the English word ânovelâ and the French word ânuovelleâ. Citation Sentence: The emphasis on narrativity takes into account the use of blogs as public diaries on the web , that is still the main current interpretation of this literary genre , or metagenre ( McNeill , 2005 ) . Context after the citation: Furthermore we noticed that blogs and wikis are currently subjected to osmosis, because they have in common the underlying core technology. So blogs are a literary metagenre which started as authored personal diaries or journals. Now they try to collect themselves in so-called âblogspheresâ. On the other side, wikis started as collective works where each entry is not owned by a single author e.g. Wikipedia (2005).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1044
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: For example, it would be helpful to consider strong correspondence between certain English and Chinese words, as in (Wu, 1994). Beside HTML markups, other criteria may also be incorporated. Despite their large length difference, the two 0002 sentences are still aligned as a 1-1 pair, because the sentences in the following 4 alignments (0003 0003; 0004 0004, 0005; 0005 0006; 0006 0007) have rather similar HTML markups and are taken by the program to be the most likely alignments. Citation Sentence: For example , it would be helpful to consider strong correspondence between certain English and Chinese words , as in ( Wu , 1994 ) . Context after the citation: We hope to implement such correspondences in our future research.	FutureWork	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1045
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: The system utilizes several large size biological databases including three NCBI databases (GenPept [11], RefSeq [12], and Entrez GENE [13]), PSD database from Protein Information Resources (PIR) [14], and Citation Sentence: The system utilizes several large size biological databases including three NCBI databases ( GenPept [ 11 ] , RefSeq [ 12 ] , and Entrez GENE [ 13 ] ) , PSD database from Protein Information Resources ( PIR ) [ 14 ] , and Context after the citation: Proceedings of the ACL Interactive Poster and Demonstration Sessions, pages 17â20, Ann Arbor, June 2005. c ï¿½2005 Association for Computational Linguistics UniProt [15]. Additionally, several model organism databases or nomenclature databases were used.	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1046
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: We take some core ideas from our previous work on mining script information (Regneri et al., 2010). However, they use MSA at the sentence level rather than at the discourse level. Our approach for sentential paraphrase extraction is related to the one introduced by Barzilay and Lee (2003), who also employ multiple sequence alignment (MSA). Citation Sentence: We take some core ideas from our previous work on mining script information ( Regneri et al. , 2010 ) . Context after the citation: In this earlier work, we focused on event structures and their possible realizations in natural language. The corpus used in those experiments were short crowd-sourced descriptions of everyday tasks written in bullet point style. We aligned them with a hand-crafted similarity measure that was specifically designed for this text type. In this current work, we target the general task of extracting paraphrases for events rather than the much more specific scriptrelated task.	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1047
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: While IA is generally thought to be consistent with findings on human language production (Hermann and Deutsch 1976; Levelt 1989; Pechmann 1989; Sonnenschein 1982), the hypothesis that incrementality is a good model of human GRE seems unfalsifiable until a preference order is specified for the properties on which it operates. Citation Sentence: While IA is generally thought to be consistent with findings on human language production ( Hermann and Deutsch 1976 ; Levelt 1989 ; Pechmann 1989 ; Sonnenschein 1982 ) , the hypothesis that incrementality is a good model of human GRE seems unfalsifiable until a preference order is specified for the properties on which it operates . Context after the citation: (Wildly redundant descriptions can result if the âwrongâ preference order are chosen.) We shall see that vague descriptions pose particular challenges to incrementality. One question emerges when the IA is combined with findings on word order and incremental interpretation. If human speakers and/or writers perform CD incrementally, then why are properties not expressed in the same order in which they were selected?	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1048
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: Typical examples are Bulgarian (Simov et al., 2005; Simov and Osenova, 2003), Chinese (Chen et al., 2003), Danish (Kromann, 2003), and Swedish (Nilsson et al., 2005). If we start by considering languages with a labeled attachment score of 85% or higher, they are characterized by high precision and recall for root nodes, typically 95/90, and by a graceful degradation of attachment score as arcs grow longer, typically 95â90â85, for arcs of length 1, 2 and 3â6. before we turn to Swedish and Turkish, focusing on recall and precision of root nodes, as a reflection of global syntactic structure, and on attachment score as a function of arc length. Citation Sentence: Typical examples are Bulgarian ( Simov et al. , 2005 ; Simov and Osenova , 2003 ) , Chinese ( Chen et al. , 2003 ) , Danish ( Kromann , 2003 ) , and Swedish ( Nilsson et al. , 2005 ) . Context after the citation: Japanese (Kawata and Bartels, 2000), despite a very high accuracy, is different in that attachment score drops from 98% to 85%, as we go from length 1 to 2, which may have something to do with the data consisting of transcribed speech with very short utterances. A second observation is that a high proportion of non-projective structures leads to fragmentation in the parser output, reflected in lower precision for roots. This is noticeable for German (Brants et al., 2002) and Portuguese (Afonso et al., 2002), which still have high overall accuracy thanks to very high attachment scores, but much more conspicuous for Czech (BÂ¨ohmovÂ´a et al., 2003), Dutch (van der Beek et al., 2002) and Slovene (DËzeroski et al., 2006), where root precision drops more drastically to about 69%, 71% and 41%, respectively, and root recall is also affected negatively. On the other hand, all three languages behave like high-accuracy languages with respect to attachment score.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1049
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: TNT refers to the HPSG parser (Torisawa et al., 2000), C++ implementation of the two-phase parsing algorithm that performs filtering with a compiled CFG (phase 1) and then executes feature unification (phase 2). In Table 2, lem refers to the LTAG parser (Sarkar et al., 2000), ANSI C implementation of the two-phase parsing algorithm that performs the head corner parsing (van Noord, 1994) without features (phase 1), and then executes feature unification (phase 2). Table 2 shows the average parsing time with the LTAG and HPSG parsers. Citation Sentence: TNT refers to the HPSG parser ( Torisawa et al. , 2000 ) , C++ implementation of the two-phase parsing algorithm that performs filtering with a compiled CFG ( phase 1 ) and then executes feature unification ( phase 2 ) . Context after the citation: Table 2 clearly shows that the HPSG parser is significantly faster than the LTAG parser. This result implies that parsing techniques for HPSG are also beneficial for LTAG 4We eliminated 32 elementary trees because the LTAG parser cannot produce correct derivation trees with them. 5These lexical items are a subset of the original XTAG English grammar distribution.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:105
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: â¢ A user study was performed, but it was either very small compared to the corpus (Carmel, Shtalhaim, and Soffer 2000; Jijkoun and de Rijke 2005), or the corpus itself was significantly smaller than ours (Feng et al. 2006; Leuski et al. 2006). â¢ Only an automatic evaluation was performed, which relied on having model responses (Berger and Mittal 2000; Berger et al. 2000). â¢ Only qualitative observations of the responses were reported (no formal evaluation was performed) (Lapalme and Kosseim 2003; Roy and Subramaniam 2006). Citation Sentence: â¢ A user study was performed , but it was either very small compared to the corpus ( Carmel , Shtalhaim , and Soffer 2000 ; Jijkoun and de Rijke 2005 ) , or the corpus itself was significantly smaller than ours ( Feng et al. 2006 ; Leuski et al. 2006 ) . Context after the citation: The representativeness of the sample size was not discussed in any of these studies. There are significant practical difficulties associated with conducting the user studies needed to produce meaningful results for our system. Firstly, the size of our corpus and the number of parameters and settings that we need to test mean that in order for a user study to be representative, a fairly large sample involving several hundreds of requestâresponse pairs would have to be used. Further, user-based evaluations of the output produced by our system require the subjects to read relatively long requestâ response e-mails, which quickly becomes tedious.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1050
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Our work on the prosodic phrase status of clause final prepositional phrases, which we discuss below, suggests the existence of a discourse-neutral phrasing that depends on syntactic constituency mediated by string adjacency and length of a potential prosodic phrase.3 Such phrasing provides us with a typical phrasing pattern analogous to the typical phrasal stress patterns examined in Liberman and Prince (1977), which "are often overwhelmed by the chiaroscuro of highlight and background in discourse, but retain the status of null-hypothesis patterns that emerge when Computational Linguistics Volume 16, Number 3, September 1990 157 J. Bachenko and E. Fitzpatrick Discourse-Neutral Prosodic Phrasing in English there is no good reason to take some other option" (p. 251). In other words, when prosodic features that reflect facts of the discourse are removed, is there a residual, neutral phrasing? Thus both of these efforts leave open the question as to whether discourse features completely determine prosodic phrasing or are a complement to some more basic set of determinants, syntactic and/or phonological. Citation Sentence: Our work on the prosodic phrase status of clause final prepositional phrases , which we discuss below , suggests the existence of a discourse-neutral phrasing that depends on syntactic constituency mediated by string adjacency and length of a potential prosodic phrase .3 Such phrasing provides us with a typical phrasing pattern analogous to the typical phrasal stress patterns examined in Liberman and Prince ( 1977 ) , which `` are often overwhelmed by the chiaroscuro of highlight and background in discourse , but retain the status of null-hypothesis patterns that emerge when Computational Linguistics Volume 16 , Number 3 , September 1990 157 J. Bachenko and E. Fitzpatrick Discourse-Neutral Prosodic Phrasing in English there is no good reason to take some other option '' ( p. 251 ) . Context after the citation: This approach to prosodic phrase boundary determination brings us closer to a framework in which phonological, syntactic, and discourse features all contribute to prosodic phrasing. The possibility of a discourse-neutral prosodic phrasing is also of import to the prosodic quality of synthetic speech, since it allows us to "get by" without a complete description of the discourse features of a given text, many of which have yet to be characterized. Interestingly, in the data we examined we found only 14 percent of the phrases to be discourse-determined. The identification of a preferred phrasing that is independent of discourse also aids us in identifying and characterizing the discourse features that impinge on prosodic phrasing.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1051
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Another line of research approaches grounded language knowledge by augmenting distributional approaches of word meaning with perceptual information (Andrews et al., 2009; Steyvers, 2010; Feng and Lapata, 2010b; Bruni et al., 2011; Silberer and Lapata, 2012; Johns and Jones, 2012; Bruni et al., 2012a; Bruni et al., 2012b; Silberer et al., 2013). Some efforts have tackled tasks such as automatic image caption generation (Feng and Lapata, 2010a; Ordonez et al., 2011), text illustration (Joshi et al., 2006), or automatic location identification of Twitter users (Eisenstein et al., 2010; Wing and Baldridge, 2011; Roller et al., 2012). Others provide automatic mappings of natural language instructions to executable actions, such as interpreting navigation directions (Chen and Mooney, 2011) or robot commands (Tellex et al., 2011; Matuszek et al., 2012). Citation Sentence: Another line of research approaches grounded language knowledge by augmenting distributional approaches of word meaning with perceptual information ( Andrews et al. , 2009 ; Steyvers , 2010 ; Feng and Lapata , 2010b ; Bruni et al. , 2011 ; Silberer and Lapata , 2012 ; Johns and Jones , 2012 ; Bruni et al. , 2012a ; Bruni et al. , 2012b ; Silberer et al. , 2013 ) . Context after the citation: Although these approaches have differed in model definition, the general goal in this line of research has been to enhance word meaning with perceptual information in order to address one of the most common criticisms of distributional semantics: that the âmeaning of words is entirely given by other wordsâ (Bruni et al., 2012b). In this paper, we explore various ways to integrate new perceptual information through novel computational modeling of this grounded knowledge into a multimodal distributional model of word meaning. The model we rely on was originally developed by Andrews et al. (2009) and is based on a generalization of Latent Dirichlet Allocation.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1052
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: When objects are compared in terms of several dimensions, these dimensions can be weighed in different ways (e.g., Rasmusen 1989). 9.3.1 Combinations of Adjectives. Citation Sentence: When objects are compared in terms of several dimensions , these dimensions can be weighed in different ways ( e.g. , Rasmusen 1989 ) . Context after the citation: Let us focus on references to an individual referent r, starting with a description that contains more than one gradable adjective. The NP the tall fat giraffe, for example, can safely refer to an element b in a situation like the one below, where b is the only element that exceeds all distractors with respect to some dimension (a different one for a than for c, as it happens) while not being exceeded by any distractors in any dimension: Cases like this would be covered if the decision-theoretic property of Pareto optimality (e.g., Feldman 1980) was used as the sole criterion: Formally, an object r E C has a Pareto-optimal combination of Values V iff there is no other x E C such that 1. ]	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1053
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Most DOP models, such as in Bod (1993), Goodman (1996), Bonnema et al. (1997), Sima'an (2000) and Collins & Duffy (2002), use a likelihood criterion in defining the best parse tree: they take (some notion of) the most likely (i.e. most probable) tree as a candidate for the best tree of a sentence. Citation Sentence: Most DOP models , such as in Bod ( 1993 ) , Goodman ( 1996 ) , Bonnema et al. ( 1997 ) , Sima'an ( 2000 ) and Collins & Duffy ( 2002 ) , use a likelihood criterion in defining the best parse tree : they take ( some notion of ) the most likely ( i.e. most probable ) tree as a candidate for the best tree of a sentence . Context after the citation: We will refer to these models as Likelihood-DOP models, but in this paper we will specifically mean by "Likelihood-DOP" the PCFG-reduction of Bod (2001) given in Section 2.2. In Bod (2000b), an alternative notion for the best parse tree was proposed based on a simplicity criterion: instead of producing the most probable tree, this model produced the tree generated by the shortest derivation with the fewest training subtrees. We will refer to this model as Simplicity-DOP. In case the shortest derivation is not unique, Bod (2000b) proposes to back off to a frequency ordering of the subtrees.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1054
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: While IA is generally thought to be consistent with findings on human language production (Hermann and Deutsch 1976; Levelt 1989; Pechmann 1989; Sonnenschein 1982), the hypothesis that incrementality is a good model of human GRE seems unfalsifiable until a preference order is specified for the properties on which it operates. Citation Sentence: While IA is generally thought to be consistent with findings on human language production ( Hermann and Deutsch 1976 ; Levelt 1989 ; Pechmann 1989 ; Sonnenschein 1982 ) , the hypothesis that incrementality is a good model of human GRE seems unfalsifiable until a preference order is specified for the properties on which it operates . Context after the citation: (Wildly redundant descriptions can result if the âwrongâ preference order are chosen.) We shall see that vague descriptions pose particular challenges to incrementality. One question emerges when the IA is combined with findings on word order and incremental interpretation. If human speakers and/or writers perform CD incrementally, then why are properties not expressed in the same order in which they were selected?	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1055
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Other molecular biology databases We also included several model organism databases or nomenclature databases in the construction of the dictionary, i.e., mouse Mouse Genome Database (MGD) [18], fly FlyBase [19], yeast Saccharomyces Genome Database (SGD) [20], rat â Rat Genome Database (RGD) [21], worm â WormBase [22], Human Nomenclature Database (HUGO) [23], Online Mendelian Inheritance in Man (OMIM) [24], and Enzyme Nomenclature Database (ECNUM) [25, 26]. The Semantic Network contains information about the types or categories (e.g., âDisease or Syndromeâ, âVirusâ) to which all META concepts have been assigned. The SPECIALIST lexicon contains syntactic information for many terms, component words, and English words, including verbs, which do not appear in the META. Citation Sentence: Other molecular biology databases We also included several model organism databases or nomenclature databases in the construction of the dictionary , i.e. , mouse Mouse Genome Database ( MGD ) [ 18 ] , fly FlyBase [ 19 ] , yeast Saccharomyces Genome Database ( SGD ) [ 20 ] , rat -- Rat Genome Database ( RGD ) [ 21 ] , worm -- WormBase [ 22 ] , Human Nomenclature Database ( HUGO ) [ 23 ] , Online Mendelian Inheritance in Man ( OMIM ) [ 24 ] , and Enzyme Nomenclature Database ( ECNUM ) [ 25 , 26 ] . Context after the citation:	Uses	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1056
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: The RenTAL system automatically converts an FB-LTAG grammar into a strongly equivalent HPSG-style grammar (Yoshinaga and Miyao, 2001). This paper describes an approach for sharing resources in various grammar formalisms such as Feature-Based Lexicalized Tree Adjoining Grammar (FB-LTAG1) (Vijay-Shanker, 1987; Vijay-Shanker and Joshi, 1988) and Head-Driven Phrase Structure Grammar (HPSG) (Pollard and Sag, 1994) by a method of grammar conversion. Citation Sentence: The RenTAL system automatically converts an FB-LTAG grammar into a strongly equivalent HPSG-style grammar ( Yoshinaga and Miyao , 2001 ) . Context after the citation: Strong equivalence means that both grammars generate exactly equivalent parse results, and that we can share the LTAG grammars and lexicons in HPSG applications. Our system can reduce considerable workload to develop a huge resource (grammars and lexicons) from scratch. Our concern is, however, not limited to the sharing of grammars and lexicons. Strongly equivalent grammars enable the sharing of ideas developed in each formalism.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1057
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: The features can be easily obtained by modifying the TAT extraction algorithm described in (Liu et al., 2006). Note that for lexicalized tree, features do not include the information of sub-trees since there is no nonterminals. Furthermore, we also build MERS models for lexicalized and unlexicalized source trees. Citation Sentence: The features can be easily obtained by modifying the TAT extraction algorithm described in ( Liu et al. , 2006 ) . Context after the citation: When a TAT is extracted from a word-aligned, source-parsed parallel sentence, we just record the contextual features and the features of the sub-trees. Then we use the toolkit implemented by Zhang (2004) to train MERS models for the ambiguous source syntactic trees separately. We set the iteration number to 100 and Gaussian prior to 1.	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1058
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Marinov and Hemming (2004) present preliminary work on the automatic extraction of subcategorization frames for Bulgarian from the BulTreeBank (Simov, Popova, and Osenova 2002). The system learns 137 subcategorization frames from 19,126 sentences for 914 verbs (those which occurred five times or more). This is achieved using three different hypothesis tests: BHT, log-likelihood ratio, and t-score. Citation Sentence: Marinov and Hemming ( 2004 ) present preliminary work on the automatic extraction of subcategorization frames for Bulgarian from the BulTreeBank ( Simov , Popova , and Osenova 2002 ) . Context after the citation: In a similar way to that of Sarkar and Zeman (2000), Marinov and Hemmingâs system collects both arguments and adjuncts. It then uses the binomial log-likelihood ratio to filter incorrect frames. The BulTreebank trees are annotated with HPSG-typed feature structure information and thus contain more detail than the dependency trees. The work done for Bulgarian is small-scale, however, as Marinov and Hemming are working with a preliminary version of the treebank with 580 sentences.	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1059
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: 1Â° The body of a plan can be an action or sequence of actions, a goal or sequence 9 Moore and Paris also note that "a generation system must maintain the kinds of information outlined by Grosz and Sidner" (Moore and Paris 1989, 203). Each matching plan's preconditions are checked; if they are currently (believed to be) true, the planner then attempts to find all instantiations of the plan's body. The planner first checks all of its top-level plans to see which have effects that match the goal. Citation Sentence: 1 Â° The body of a plan can be an action or sequence of actions , a goal or sequence 9 Moore and Paris also note that `` a generation system must maintain the kinds of information outlined by Grosz and Sidner '' ( Moore and Paris 1989 , 203 ) . Context after the citation: Their planner uses plan structures similar to IGEN's, except that the plan operators they use are generally instantiations of rhetorical relations drawn from Rhetorical Structure Theory (Mann and Thompson 1987). In IGEN, the plans can involve any goals or actions that could be achieved via communication. Hovy has described another text planner that builds similar plans (Hovy 1988b). This system, however, starts with a list of information to be expressed and merely arranges it into a coherent pattern; it is thus not a planner in the sense used here (as Hovy makes clear).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:106
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: The use of the web as a corpus for teaching and research on language has been proposed a number of times (Kilgarriff, 2001; Robb, 2003; Rundell, 2000; Fletcher, 2001, 2004b) and received a special issue of the journal Computational Linguistics (Kilgarriff and Grefenstette, 2003). This corpus annotation bottleneck becomes even more problematic for voluminous data sets drawn from the web. Larger systems to support multiple document tagging processes would require resources that cannot be realistically provided by existing single-server systems. Citation Sentence: The use of the web as a corpus for teaching and research on language has been proposed a number of times ( Kilgarriff , 2001 ; Robb , 2003 ; Rundell , 2000 ; Fletcher , 2001 , 2004b ) and received a special issue of the journal Computational Linguistics ( Kilgarriff and Grefenstette , 2003 ) . Context after the citation: Studies have used several different methods to mine web data. Turney (2001) extracts word co-occurrence probabilities from unlabelled text collected from a web crawler. Baroni and Bernardini (2004) built a corpus by iteratively searching Google for a small set of seed terms. Prototypes of Internet search engines for linguists, corpus linguists and lexicographers have been proposed: WebCorp (Kehoe and Renouf, 2002), KWiCFinder (Fletcher, 2004a) and the Linguistâs Search Engine (Kilgarriff, 2003; Resnik and Elkiss, 2003).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1060
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: method Context before the citation: ASARES is based on a Machine Learning technique, Inductive Logic Programming (ILP) (Muggleton and De-Raedt, 1994), which infers general morpho-syntactic patterns from a set of examples (this set is noted E+ hereafter) and counter-examples (Eâ) of the elements one We simply give a short account of its basic principles herein. ASARES is presented in detail in (Claveau et al., 2003). Citation Sentence: ASARES is based on a Machine Learning technique , Inductive Logic Programming ( ILP ) ( Muggleton and De-Raedt , 1994 ) , which infers general morpho-syntactic patterns from a set of examples ( this set is noted E + hereafter ) and counter-examples ( E â ) of the elements one Context after the citation: 2CORDIAL is a commercial product of SynapseDeveloppement. wants to acquire and their context. The contextual patterns produced can then be applied to the corpus in order to retrieve new elements. The acquisition process can be summarized in 3 steps:	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1061
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Baroni and Bernardini (2004) built a corpus by iteratively searching Google for a small set of seed terms. Turney (2001) extracts word co-occurrence probabilities from unlabelled text collected from a web crawler. Studies have used several different methods to mine web data. Citation Sentence: Baroni and Bernardini ( 2004 ) built a corpus by iteratively searching Google for a small set of seed terms . Context after the citation: Prototypes of Internet search engines for linguists, corpus linguists and lexicographers have been proposed: WebCorp (Kehoe and Renouf, 2002), KWiCFinder (Fletcher, 2004a) and the Linguistâs Search Engine (Kilgarriff, 2003; Resnik and Elkiss, 2003). A key concern in corpus linguistics and related disciplines is verifiability and replicability of the results of studies. Word frequency counts in internet search engines are inconsistent and unreliable (Veronis, 2005). Tools based on static corpora do not suffer from this problem, e.g. BNCweb7, developed at the University of Zurich, and View 8 (Variation in English Words and Phrases, developed at Brigham Young University)	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1062
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: None Context before the citation: Expanding on a suggestion of Michiels (1982), we classify verbs as Subject Equi, Object Equi, Subject Raising or Object Raising for each sense which has a predicate complement code associated with it. In these situations the parser can use the semantic type of the verb to compute this relationship. In any subcategorisation frame which involves a predicate complement there will be a non-transparent relationship between the superficial syntactic form and the underlying logical relations in the sentence. Citation Sentence: Expanding on a suggestion of Michiels ( 1982 ) , we classify verbs as Subject Equi , Object Equi , Subject Raising or Object Raising for each sense which has a predicate complement code associated with it . Context after the citation: These terms, which derive from Transformational Grammar, are used as convenient labels for what we regard as a semantic distinction; the actual output of the program is a specification of the mapping from superficial syntactic form to an underlying logical representation. For example, labelling believe(3) (Type 2 ORaising) indicates that this is a two place predicate and that, if believe(3) occurs with a syntactic direct object, as in (1) John believes the Earth to be round it will function as the logical subject of the predicate complement.	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1063
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: This Principle of Finitism is also assumed by Johnson-Laird (1983), Jackendoff (1983), Kamp (1981), and implicitly or explicitly by almost all researchers in computational linguistics. The issues of control are not so important for us at this point; we restrict ourselves to describing the logic. Therefore these notions, and all other constructs we are going to define (axioms, metarules, definitions etc.) are computational, although usually we will not provide explicit algorithms for computing them. Citation Sentence: This Principle of Finitism is also assumed by Johnson-Laird ( 1983 ) , Jackendoff ( 1983 ) , Kamp ( 1981 ) , and implicitly or explicitly by almost all researchers in computational linguistics . Context after the citation: As a logical postulate it is not very radical; it is possible within a finitary framework to develop that part of mathematics that is used or has potential applications in natural science, such as mathematical analysis (cfXXX Mycielski 1981). On the other hand, a possible obstacle to our strategy of using only finite objects is the fact that the deductive closure of any set of formulas is not finite in standard logic, while, clearly, we will have to deduce new facts from formal representations of text and background knowledge. But there are several ways to avoid this obstruction. For example, consider theories consisting of universal formulas without function symbols.	CompareOrContrast	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1064
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: experiments Context before the citation: For more information on CATiB, see Habash and Roth (2009) and Habash, Faraj, and Roth (2009). For the corpus statistics, see Table 1. An example CATiB dependency tree is shown in Figure 1. Citation Sentence: For more information on CATiB , see Habash and Roth ( 2009 ) and Habash , Faraj , and Roth ( 2009 ) . Context after the citation:	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1065
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: With the exception of (Fung, 1995b), previous methods for automatically constructing statistical translation models begin by looking at word cooccurrence frequencies in bitexts (Gale & Church, 1991; Kumano & Hirakawa, 1994; Fung, 1995a; Melamed, 1995). 2 Co-occurrence The hidden parameters can be conditioned on prior knowledge about the bitext to improve the model's accuracy. Citation Sentence: With the exception of ( Fung , 1995b ) , previous methods for automatically constructing statistical translation models begin by looking at word cooccurrence frequencies in bitexts ( Gale & Church , 1991 ; Kumano & Hirakawa , 1994 ; Fung , 1995a ; Melamed , 1995 ) . Context after the citation: A bitext comprises a pair of texts in two languages, where each text is a translation of the other. Word co-occurrence can be defined in various ways. The most common way is to divide each half of the bitext into an equal number of segments and to align the segments so that each pair of segments Si and Ti are translations of each other (Gale & Church, 1991; Melamed, 1996a). Then, two word tokens (u, v) are said to co-occur in the	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1066
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: introduction Context before the citation: Experiments (Section 5) show that forestbased extraction improves BLEU score by over 1 point on a state-of-the-art tree-to-string system (Liu et al., 2006; Mi et al., 2008), which is also 0.5 points better than (and twice as fast as) extracting on 30-best parses. We instead propose a novel approach that extracts rules from packed forests (Section 3), which compactly encodes many more alternatives than kbest lists. also inefficient to extract rules separately from each of these very similar trees (or from the cross-product of k2 similar tree-pairs in tree-to-tree models). Citation Sentence: Experiments ( Section 5 ) show that forestbased extraction improves BLEU score by over 1 point on a state-of-the-art tree-to-string system ( Liu et al. , 2006 ; Mi et al. , 2008 ) , which is also 0.5 points better than ( and twice as fast as ) extracting on 30-best parses . Context after the citation: When combined with our previous orthogonal work on forest-based decoding (Mi et al., 2008), the forest-forest approach achieves a 2.5 BLEU points improvement over the baseline, and even outperforms the hierarchical system of Hiero, one of the best-performing systems to date. Besides tree-to-string systems, our method is also applicable to other paradigms such as the string-totree models (Galley et al., 2006) where the rules are in the reverse order, and easily generalizable to pairs of forests in tree-to-tree models. 2 Tree-based Translation We review in this section the tree-based approach to machine translation (Liu et al., 2006; Huang et al., 2006), and its rule extraction algorithm (Galley et al., 2004; Galley et al., 2006).	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1067
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Our approach to extract and classify social events builds on our previous work (Agarwal and Rambow, 2010), which in turn builds on work from the relation extraction community (Nguyen et al., 2009). Take your own.â (6) âTake it,â said Emma, smiling, and pushing the paper towards Harriet âit is for you. Citation Sentence: Our approach to extract and classify social events builds on our previous work ( Agarwal and Rambow , 2010 ) , which in turn builds on work from the relation extraction community ( Nguyen et al. , 2009 ) . Context after the citation: Therefore, the task of relation extraction is most closely related to the tasks addressed in this paper. Researchers have used other notions of semantics in the literature such as latent semantic analysis (Plank and Moschitti, 2013) and relation-specific semantics (Zelenko et al., 2003; Culotta and Sorensen, 2004). To the best of our knowledge, there is only one work that uses frame semantics for relation extraction (Harabagiu et al., 2005). Harabagiu et al. (2005) propose a novel semantic kernel that incorporates frame parse information in the kernel computation that calculates similarity between two dependency trees.	Extends	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1068
You will be presented with a citation segment from the section of an NLP research paper, as well as the context surrounding that citation. Classify the intent behind this citation by choosing from one of the following categories: - Background: provides context or foundational information related to the topic. - Extends: builds upon the cited work. - Uses: applies the methods or findings of the cited work. - Motivation: cites the work as inspiration or rationale for the research. - CompareOrContrast: compares or contrasts the cited work with others. - FutureWork: cites the work as a direction for future research. Your answer should be a single word from the following list of options: ["Background", "Extends", "Uses", "Motivation", "CompareOrContrast", "FutureWork"]. Do not include any other text in your response. Section Title: related work Context before the citation: Previous sentiment-analysis work in different domains has considered inter-document similarity (Agarwal and Bhattacharyya, 2005; Pang and Lee, 2005; Goldberg and Zhu, 2006) or explicit Relationships between the unlabeled items Carvalho and Cohen (2005) consider sequential relations between different types of emails (e.g., between requests and satisfactions thereof) to classify messages, and thus also explicitly exploit the structure of conversations. We currently do not have an efficient means to encode disagreement information as hard constraints; we plan to investigate incorporating such information in future work. Citation Sentence: Previous sentiment-analysis work in different domains has considered inter-document similarity ( Agarwal and Bhattacharyya , 2005 ; Pang and Lee , 2005 ; Goldberg and Zhu , 2006 ) or explicit Context after the citation: inter-document references in the form of hyperlinks (Agrawal et al., 2003). Notable early papers on graph-based semisupervised learning include Blum and Chawla (2001), Bansal et al. (2002), Kondor and Lafferty (2002), and Joachims (2003). Zhu (2005) maintains a survey of this area. Recently, several alternative, often quite sophisticated approaches to collective classification have been proposed (Neville and Jensen, 2000; Lafferty et al., 2001; Getoor et al., 2002; Taskar et al., 2002; Taskar et al., 2003; Taskar et al., 2004; McCallum and Wellner, 2004).	Background	{ "domains": [ "artificial_intelligence" ], "input_context": "multiple_paragraphs", "output_context": "label", "source_type": "single_source", "task_family": "classification" }	acl_arc_intent_classification:train:1069

End of preview. Expand in Dataset Viewer.

SciRIFF

The SciRIFF dataset includes 137K instruction-following demonstrations for 54 scientific literature understanding tasks. The tasks cover five essential scientific literature categories and span five domains. The dataset is described in our paper SciRIFF: A Resource to Enhance Language Model Instruction-Following over Scientific Literature.

There are three dataset configurations with different max context lengths: 4096, 8192, and 16384. All experiments in the paper are performed with the 4096 context window. You can load the dataset like:

import datasets
ds = datasets.load_dataset("allenai/SciRIFF", "4096")

Code to create the dataset, train models on SciRIFF, and perform evaluation is available at our GitHub repo: https://github.com/allenai/SciRIFF. To train models on SciRIFF data, you should use the SciRIFF train mix dataset.

Table of Contents

Dataset details
License
Task provenance
Task metadata

Dataset details

Each instance in SciRIFF has the following fields:

input: Task input (i.e. user message).
output: Task output (i.e. expected model response).
_instance_id: A unique id for the instance, formatted like {task_name}:{split}:{instance_id}. For instance, qasa_abstractive_qa:test:182.
metadata: Task metadata. More information on the schema for task metadata can be found in the SciRIFF GitHub repo.
- task_family: The category to which this task belongs. Options include summarization, ie, qa, entailment, and classification. Some categories have sub-categories which are largely self-explanatory; see the repo for more information.
- domains: Scientific field(s) that the task covers. Options include: clinical_medicine, biomedicine, chemistry, artificial_intelligence, materials_science, and misc.
- input_context: Whether the input is a paragraph, full text, etc. Options include: sentence, paragraph, multiple_paragraphs (including full paper text), and structured (e.g. code for a LaTex table).
- source_type: Indicates whether the input comes from a single paper or multiple. Options include single_source, multiple_source.
- output_context: Options include: label, sentence, paragraph, multiple_paragraphs, json, jsonlines.

License

SciRIFF is licensed under ODC-By. Licenses of the datasets from which SciRIFF is derived are listed below.

Task provenance

SciRIFF was created by repurposing existing scientific literature understanding datasets. Below we provide information on the source data for each SciRIFF task, including license information on individual datasets where available. Where possible, we leveraged the BigBIO collection as a starting point, rather than reprocessing datasets from scratch. In the table below, we include the name of the BigBio subset for all tasks available in BigBio; these can be loaded like datasets.load_dataset(bigbio/{bigbio_subset}).

SciRIFF Name	Paper Link	License	Website / Download Link	BigBio Subset
`acl_arc_intent_classification`	ACL ARC	-	https://github.com/allenai/scicite/
`anat_em_ner`	AnatEM	CC BY	https://nactem.ac.uk/anatomytagger/#AnatEM	`anat_em`
`annotated_materials_syntheses_events`	Materials Science Procedural Text Corpus	MIT	https://github.com/olivettigroup/annotated-materials-syntheses
`bc7_litcovid_topic_classification`	BioCreative VII LitCOVID	-	https://biocreative.bioinformatics.udel.edu/tasks/biocreative-vii/track-5/	`bc7_litcovid`
`bioasq_{factoid,general,list,yesno}_qa`	BioASQ	CC BY	http://bioasq.org/	`bioasq`
`biored_ner`	BioRED	-	https://ftp.ncbi.nlm.nih.gov/pub/lu/BioRED/	`biored`
`cdr_ner`	BioCreative V CDR	-	https://biocreative.bioinformatics.udel.edu/tasks/biocreative-v/track-3-cdr/	`bc5cdr`
`chemdner_ner`	CHEMDNER	-	https://biocreative.bioinformatics.udel.edu/resources/biocreative-iv/chemdner-corpus/	`chemdner`
`chemprot_{ner,re}`	BioCreative VI ChemProt	-	https://biocreative.bioinformatics.udel.edu/news/corpora/chemprot-corpus-biocreative-vi/	`chemprot`
`chemsum_single_document_summarization`	ChemSum	-	https://github.com/griff4692/calibrating-summaries
`chemtables_te`	ChemTables	GPL 3.0	https://huggingface.co/datasets/fbaigt/schema-to-json
`chia_ner`	Chia	CC BY	https://github.com/WengLab-InformaticsResearch/CHIA	`chia`
`covid_deepset_qa`	COVID-QA	Apache 2.0	https://github.com/deepset-ai/COVID-QA	`covid_qa_deepset`
`covidfact_entailment`	CovidFact	-	https://github.com/asaakyan/covidfact
`craftchem_ner`	CRAFT-Chem	-	https://huggingface.co/datasets/ghadeermobasher/CRAFT-Chem
`data_reco_mcq_{mc,sc}`	DataFinder	Apache 2.0	https://github.com/viswavi/datafinder/tree/main
`ddi_ner`	DDI	CC BY	https://github.com/isegura/DDICorpus	`ddi_corpus`
`discomat_te`	DISCoMaT	CC BY-SA	https://github.com/M3RG-IITD/DiSCoMaT
`drug_combo_extraction_re`	Drug Combinations	-	https://github.com/allenai/drug-combo-extraction
`evidence_inference`	Evidence inference	MIT	https://evidence-inference.ebm-nlp.com/
`genia_ner`	JNLPBA	CC BY	https://github.com/spyysalo/jnlpba	`jnlpba`
`gnormplus_ner`	GNormPlus	-	https://www.ncbi.nlm.nih.gov/research/bionlp/Tools/gnormplus/	`gnormplus`
`healthver_entailment`	HealthVer	nan	https://github.com/sarrouti/healthver
`linnaeus_ner`	LINNAEUS	CC BY	https://sourceforge.net/projects/linnaeus/	`linnaeus`
`medmentions_ner`	MedMentions	CC 0	https://github.com/chanzuckerberg/MedMentions	`medmentions`
`mltables_te`	AxCell	Apache 2.0	https://github.com/paperswithcode/axcell
`mslr2022_cochrane_multidoc_summarization`	Cochrane	Apache 2.0	https://github.com/allenai/mslr-shared-task
`mslr2022_ms2_multidoc_summarization`	MS^2	Apache 2.0	https://github.com/allenai/mslr-shared-task
`multicite_intent_classification`	MultiCite	CC BY-NC	https://github.com/allenai/multicite
`multixscience_multidoc_summarization`	Multi-XScience	MIT	https://github.com/yaolu/Multi-XScience
`mup_single_document_summarization`	MUP	Apache 2.0	https://github.com/allenai/mup
`ncbi_ner`	NCBI Disease	CC 0	https://www.ncbi.nlm.nih.gov/CBBresearch/Dogan/DISEASE/	`ncbi_disease`
`nlmchem_ner`	NLM-Chem	CC 0	https://ftp.ncbi.nlm.nih.gov/pub/lu/BC7-NLM-Chem-track/	`nlmchem`
`nlmgene_ner`	NLM-Gene	CC 0	https://ftp.ncbi.nlm.nih.gov/pub/lu/NLMGene/	`nlm_gene`
`pico_ner`	EBM-NLP PICO	-	https://github.com/bepnye/EBM-NLP	`pico_extraction`
`pubmedqa_qa`	PubMedQA	MIT	https://github.com/pubmedqa/pubmedqa	`pubmed_qa`
`qasa_abstractive_qa`	QASA	MIT	https://github.com/lgresearch/QASA
`qasper_{abstractive,extractive}_qa`	Qasper	CC BY	https://allenai.org/data/qasper
`scicite_classification`	SciCite	-	https://allenai.org/data/scicite
`scientific_lay_summarisation_` `{elife,plos}_single_doc_summ`	Lay Summarisation	-	https://github.com/TGoldsack1/Corpora_for_Lay_Summarisation
`scientific_papers_summarization_` `single_doc_{arxiv,pubmed}`	Scientific Papers	-	https://huggingface.co/datasets/armanc/scientific_papers
`scierc_{ner,re}`	SciERC	-	http://nlp.cs.washington.edu/sciIE/
`scifact_entailment`	SciFact	CC BY-NC	https://allenai.org/data/scifact
`scireviewgen_multidoc_summarization`	SciReviewGen	CC BY-NC	https://github.com/tetsu9923/SciReviewGen
`scitldr_aic`	SciTLDR	Apache 2.0	https://github.com/allenai/scitldr

Task metadata

Below we include metadata on each task, as described in the metadata fields above.

SciRIFF Name	Task Family	Domains	Input Context	Source Type	Output Context
`acl_arc_intent_classification`	classification	artificial_intelligence	multiple_paragraphs	single_source	label
`anat_em_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`annotated_materials_syntheses_events`	ie.event_extraction	materials_science	paragraph	single_source	json
`bc7_litcovid_topic_classification`	classification	clinical_medicine	paragraph	single_source	json
`bioasq_factoid_qa`	qa.abstractive	biomedicine	multiple_paragraphs	multiple_source	sentence
`bioasq_general_qa`	qa.abstractive	biomedicine	multiple_paragraphs	multiple_source	sentence
`bioasq_list_qa`	qa.abstractive	biomedicine	multiple_paragraphs	multiple_source	json
`bioasq_yesno_qa`	qa.yes_no	biomedicine	multiple_paragraphs	multiple_source	label
`biored_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`cdr_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`chemdner_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`chemprot_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`chemprot_re`	ie.relation_extraction	biomedicine	paragraph	single_source	json
`chemsum_single_document_summarization`	summarization	chemistry	multiple_paragraphs	single_source	paragraph
`chemtables_te`	ie.structure_to_json	chemistry	structured	single_source	jsonlines
`chia_ner`	ie.named_entity_recognition	clinical_medicine	paragraph	single_source	json
`covid_deepset_qa`	qa.extractive	biomedicine	paragraph	single_source	sentence
`covidfact_entailment`	entailment	biomedicine, clinical_medicine	paragraph	single_source	json
`craftchem_ner`	ie.named_entity_recognition	biomedicine	sentence	single_source	json
`data_reco_mcq_mc`	qa.multiple_choice	artificial_intelligence	multiple_paragraphs	multiple_source	json
`data_reco_mcq_sc`	qa.multiple_choice	artificial_intelligence	multiple_paragraphs	multiple_source	label
`ddi_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`discomat_te`	ie.structure_to_json	materials_science	structured	single_source	jsonlines
`drug_combo_extraction_re`	ie.relation_extraction	clinical_medicine	paragraph	single_source	json
`evidence_inference`	ie.relation_extraction	clinical_medicine	paragraph	single_source	json
`genia_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`gnormplus_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`healthver_entailment`	entailment	clinical_medicine	paragraph	single_source	json
`linnaeus_ner`	ie.named_entity_recognition	biomedicine	multiple_paragraphs	single_source	json
`medmentions_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`mltables_te`	ie.structure_to_json	artificial_intelligence	structured	single_source	jsonlines
`mslr2022_cochrane_multidoc_summarization`	summarization	clinical_medicine	paragraph	multiple_source	paragraph
`mslr2022_ms2_multidoc_summarization`	summarization	clinical_medicine	paragraph	multiple_source	paragraph
`multicite_intent_classification`	classification	artificial_intelligence	paragraph	single_source	json
`multixscience_multidoc_summarization`	summarization	artificial_intelligence, biomedicine, materials_science, misc	multiple_paragraphs	multiple_source	paragraph
`mup_single_document_summarization`	summarization	artificial_intelligence	multiple_paragraphs	single_source	paragraph
`ncbi_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`nlmchem_ner`	ie.named_entity_recognition	biomedicine	multiple_paragraphs	single_source	json
`nlmgene_ner`	ie.named_entity_recognition	biomedicine	paragraph	single_source	json
`pico_ner`	ie.named_entity_recognition	clinical_medicine	paragraph	single_source	json
`pubmedqa_qa`	qa.yes_no	biomedicine	paragraph	single_source	label
`qasa_abstractive_qa`	qa.abstractive	artificial_intelligence	multiple_paragraphs	single_source	paragraph
`qasper_abstractive_qa`	qa.abstractive	artificial_intelligence	multiple_paragraphs	single_source	json
`qasper_extractive_qa`	qa.extractive	artificial_intelligence	multiple_paragraphs	single_source	json
`scicite_classification`	classification	artificial_intelligence	paragraph	single_source	label
`scientific_lay_summarisation_` `elife_single_doc_summ`	summarization	biomedicine	multiple_paragraphs	single_source	paragraph
`scientific_lay_summarisation_` `plos_single_doc_summ`	summarization	biomedicine	multiple_paragraphs	single_source	paragraph
`scientific_papers_summarization_single_doc_arxiv`	summarization	artificial_intelligence, misc	multiple_paragraphs	single_source	paragraph
`scientific_papers_summarization_single_doc_pubmed`	summarization	biomedicine	multiple_paragraphs	single_source	paragraph
`scierc_ner`	ie.named_entity_recognition	artificial_intelligence	paragraph	single_source	json
`scierc_re`	ie.relation_extraction	artificial_intelligence	paragraph	single_source	json
`scifact_entailment`	entailment	biomedicine, clinical_medicine	paragraph	single_source	json
`scireviewgen_multidoc_summarization`	summarization	artificial_intelligence	multiple_paragraphs	multiple_source	paragraph
`scitldr_aic`	summarization	artificial_intelligence	multiple_paragraphs	single_source	sentence

Downloads last month: 365

Edit dataset card

Size of downloaded dataset files:

1.42 GB

Size of the auto-converted Parquet files:

1.42 GB

Number of rows:

432,740

Collection including allenai/SciRIFF

SciRIFF

Collection

Data and models to enhance instruction-following for scientific literature understanding. • 9 items • Updated 8 days ago • 8