DAGN: Discourse-Aware Graph Network for Logical Reasoning

Yinya Huang, Meng Fang, Yu Cao, Liwei Wang, Xiaodan Liang


Abstract
Recent QA with logical reasoning questions requires passage-level relations among the sentences. However, current approaches still focus on sentence-level relations interacting among tokens. In this work, we explore aggregating passage-level clues for solving logical reasoning QA by using discourse-based information. We propose a discourse-aware graph network (DAGN) that reasons relying on the discourse structure of the texts. The model encodes discourse information as a graph with elementary discourse units (EDUs) and discourse relations, and learns the discourse-aware features via a graph network for downstream QA tasks. Experiments are conducted on two logical reasoning QA datasets, ReClor and LogiQA, and our proposed DAGN achieves competitive results. The source code is available at https://github.com/Eleanor-H/DAGN.
Anthology ID:
2021.naacl-main.467
Volume:
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Month:
June
Year:
2021
Address:
Online
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5848–5855
Language:
URL:
https://aclanthology.org/2021.naacl-main.467
DOI:
10.18653/v1/2021.naacl-main.467
Bibkey:
Cite (ACL):
Yinya Huang, Meng Fang, Yu Cao, Liwei Wang, and Xiaodan Liang. 2021. DAGN: Discourse-Aware Graph Network for Logical Reasoning. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 5848–5855, Online. Association for Computational Linguistics.
Cite (Informal):
DAGN: Discourse-Aware Graph Network for Logical Reasoning (Huang et al., NAACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.naacl-main.467.pdf
Video:
 https://aclanthology.org/2021.naacl-main.467.mp4
Code
 Eleanor-H/DAGN +  additional community code
Data
DROPHotpotQALogiQAReClor