Please use this identifier to cite or link to this item:
DC FieldValue
dc.titleZero-shot Fact Verification by Claim Generation
dc.contributor.authorPan, L
dc.contributor.authorChen, W
dc.contributor.authorXiong, W
dc.contributor.authorKan, MY
dc.contributor.authorWang, WY
dc.identifier.citationPan, L, Chen, W, Xiong, W, Kan, MY, Wang, WY (2021-01-01). Zero-shot Fact Verification by Claim Generation 2 : 476-483. ScholarBank@NUS Repository.
dc.description.abstractNeural models for automated fact verification have achieved promising results thanks to the availability of large, human-annotated datasets. However, for each new domain that requires fact verification, creating a dataset by manually writing claims and linking them to their supporting evidence is expensive. We develop QACG, a framework for training a robust fact verification model by using automaticallygenerated claims that can be supported, refuted, or unverifiable from evidence from Wikipedia. QACG generates question-answer pairs from the evidence and then convert them into different types of claims. Experiments on the FEVER dataset show that our QACG framework significantly reduces the demand for human-annotated training data. In a zeroshot scenario, QACG improves a RoBERTa model's F1 from 50% to 77%, equivalent in performance to 2K+ manually-curated examples. Our QACG code is publicly available.
dc.typeConference Paper
dc.contributor.departmentDEPARTMENT OF COMPUTER SCIENCE
Appears in Collections:Staff Publications

Show simple item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
2105.14682.pdf777.89 kBAdobe PDF



Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.