Show simple item record

dc.contributor.authorThorne, J
dc.contributor.authorVlachos, Andreas
dc.date.accessioned2021-11-25T11:49:36Z
dc.date.available2021-11-25T11:49:36Z
dc.date.issued2021
dc.identifier.isbn9781954085527
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/331131
dc.description.abstractThis paper introduces the task of factual error correction: performing edits to a claim so that the generated rewrite is better supported by evidence. This extends the well-studied task of fact verification by providing a mechanism to correct written texts that are refuted or only partially supported by evidence. We demonstrate that it is feasible to train factual error correction systems from existing fact checking datasets which only contain labeled claims accompanied by evidence, but not the correction. We achieve this by employing a two-stage distant supervision approach that incorporates evidence into masked claims when generating corrections. Our approach, based on the T5 transformer and using retrieved evidence, achieved better results than existing work which used a pointer copy network and gold evidence, producing accurate factual error corrections for 5x more instances in human evaluation and a.125 increase in SARI score. The evaluation is conducted on a dataset of 65,000 instances based on a recent fact verification shared task and we release it to enable further work on the task.
dc.publisherAssociation for Computational Linguistics
dc.rightsAttribution 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.titleEvidence-based factual error correction
dc.typeConference Object
prism.endingPage3309
prism.publicationDate2021
prism.publicationNameACL-IJCNLP 2021 - 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Proceedings of the Conference
prism.startingPage3298
dc.identifier.doi10.17863/CAM.78578
dc.identifier.doi10.17863/CAM.78578
dcterms.dateAccepted2021-05-05
rioxxterms.versionofrecord10.18653/v1/2021.acl-long.256
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.licenseref.startdate2021-01-01
dc.contributor.orcidVlachos, Andreas [0000-0003-2123-5071]
dc.publisher.urlhttps://aclanthology.org/2021.acl-long.256
rioxxterms.typeConference Paper/Proceeding/Abstract
pubs.funder-project-idEuropean Commission Horizon 2020 (H2020) ERC (865958)
pubs.conference-nameProceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)
pubs.conference-start-date2021-08
pubs.conference-finish-date2021-08


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution 4.0 International
Except where otherwise noted, this item's licence is described as Attribution 4.0 International