The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.
Dataset Card for CANLI
Dataset Summary
The disambiguation of causative-passive homonymy (CPH) is potentially tricky for machines, as the causative and the passive are not distinguished by the sentences syntactic structure. By transforming CPH disambiguation to a challenging natural language inference (NLI) task, we present the first Chinese Adversarial NLI challenge set (CANLI). We show that the pretrained transformer model RoBERTa, fine-tuned on an existing large-scale Chinese NLI benchmark dataset, performs poorly on CANLI. We also employ Word Sense Disambiguation as a probing task to investigate to what extent the CPH feature is captured in the models internal representation. We find that the models performance on CANLI does not correspond to its internal representation of CPH, which is the crucial linguistic ability central to the CANLI dataset.
Languages
Chinese Mandarin
Citation Information
@inproceedings{xu-markert-2022-chinese,
title = "The {C}hinese Causative-Passive Homonymy Disambiguation: an adversarial Dataset for {NLI} and a Probing Task",
author = "Xu, Shanshan and Markert, Katja",
booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
month = jun,
year = "2022",
address = "Marseille, France",
publisher = "European Language Resources Association",
url = "https://aclanthology.org/2022.lrec-1.460",
pages = "4316--4323",
}
- Downloads last month
- 40