DAGN: Discourse-Aware Graph Network for Logical Reasoning

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 5848–5855 June 6–11, 2021. ©2021 Association for Computational Linguistics 5848 DAGN: Discourse-Aware Graph Network for Logical Reasoning Yinya Huang1∗ Meng Fang2 Yu Cao3 Liwei Wang4 Xiaodan Liang1† 1Shenzhen Campus of Sun Yat-sen University 2Tencent Robotics X 3School of Computer Science, The University of Sydney 4The Chinese University of Hong Kong yinya.huang@hotmail, mfang@tencent.com, ycao8647@uni.sydney.edu.au, lwwang@cse.cuhk.edu.hk, xdliang328@gmail.com Abstract Recent QA with logical reasoning questions re- quires passage-level relations among the sen- tences. However, current approaches still focus on sentence-level relations interacting among tokens. In this work, we explore aggre- gating passage-level clues for solving logical reasoning QA by using discourse-based infor- mation. We propose a discourse-aware graph network (DAGN) that reasons relying on the discourse structure of the texts. The model en- codes discourse information as a graph with elementary discourse units (EDUs) and dis- course relations, and learns the discourse- aware features via a graph network for down- stream QA tasks. Experiments are conducted on two logical reasoning QA datasets, Re- Clor and LogiQA, and our proposed DAGN achieves competitive results. The source code is available at https://github.com/Eleanor- H/DAGN. 1 Introduction A variety of QA datasets have promoted the devel- opment of reading comprehensions, for instance, SQuAD (Rajpurkar et al., 2016), HotpotQA (Yang et al., 2018), DROP (Dua et al., 2019), and so on. Recently, QA datasets with more complicated reasoning types, i.e., logical reasoning, are also introduced, such as ReClor (Yu et al., 2020) and LogiQA (Liu et al., 2020). The logical questions are taken from standardized exams such as GMAT and LSAT, and require QA models to read compli- cated argument passages and identify logical rela- tionships therein. For example, selecting a correct assumption that supports an argument, or finding out a claim that weakens an argument in a passage. Such logical reasoning is beyond the capability of most of the previous QA models which focus on reasoning with entities or numerical keywords. ∗This work was done during Yinya Huang’s internship in Tencent with L. Wang and M. Fang. †Corresponding Author: Xiaodan Liang. A main challenge for the QA models is to un- cover the logical structures under passages, such as identifying claims or hypotheses, or pointing out flaws in arguments. To achieve this, the QA models should first be aware of logical units, which can be sentences or clauses or other meaningful text spans, then identify the logical relationships between the units. However, the logical structures are usually hidden and difficult to be extracted, and most datasets do not provide such logical structure annotations. An intuitive idea for unwrapping such logical in- formation is using discourse relations. For instance, as a conjunction, “because” indicates a causal re- lationship, whereas “if” indicates a hypothetical relationship. However, such discourse-based infor- mation is seldom considered in logical reasoning tasks. Modeling logical structures is still lacking in logical reasoning tasks, while current opened meth- ods use contextual pre-trained models (Yu et al., …

Content truncated to 500 words