Open-Domain Conversational Question Answering with Historical Answers
Journal
2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing - Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022
ISBN
9781959429043
Date Issued
2022-01-01
Author(s)
Abstract
Open-domain conversational question answering can be viewed as two tasks: passage retrieval and conversational question answering,where the former relies on selecting candidatepassages from a large corpus and the latter requires better understanding of a question withcontexts to predict the answers. This paperproposes ConvADR-QA that leverages historical answers to boost retrieval performance andfurther achieves better answering performance.Our experiments on the benchmark dataset,OR-QuAC, demonstrate that our model outperforms existing baselines in both extractiveand generative reader settings, well justifyingthe effectiveness of historical answers for opendomain conversational question answering.
Type
conference paper