Transformers to Learn Hierarchical Contexts in Multiparty Dialogue for Span-based Question Answering

Changmao Li, Jinho D. Choi


Abstract
We introduce a novel approach to transformers that learns hierarchical representations in multiparty dialogue. First, three language modeling tasks are used to pre-train the transformers, token- and utterance-level language modeling and utterance order prediction, that learn both token and utterance embeddings for better understanding in dialogue contexts. Then, multi-task learning between the utterance prediction and the token span prediction is applied to fine-tune for span-based question answering (QA). Our approach is evaluated on the FriendsQA dataset and shows improvements of 3.8% and 1.4% over the two state-of-the-art transformer models, BERT and RoBERTa, respectively.
Anthology ID:
2020.acl-main.505
Volume:
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
Month:
July
Year:
2020
Address:
Online
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5709–5714
Language:
URL:
https://aclanthology.org/2020.acl-main.505
DOI:
10.18653/v1/2020.acl-main.505
Bibkey:
Cite (ACL):
Changmao Li and Jinho D. Choi. 2020. Transformers to Learn Hierarchical Contexts in Multiparty Dialogue for Span-based Question Answering. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 5709–5714, Online. Association for Computational Linguistics.
Cite (Informal):
Transformers to Learn Hierarchical Contexts in Multiparty Dialogue for Span-based Question Answering (Li & Choi, ACL 2020)
Copy Citation:
PDF:
https://preview.aclanthology.org/starsem-semeval-split/2020.acl-main.505.pdf
Supplementary:
 2020.acl-main.505.Supplementary.pdf
Video:
 http://slideslive.com/38928874