Plug-and-Play Module for Commonsense Reasoning in Machine Reading Comprehension

Damai Dai,Hua Zheng,Zhifang Sui,Baobao Chang

DOI: https://doi.org/10.1007/978-3-031-17189-5_3

2022-01-01

Abstract:Conventional Machine Reading Comprehension (MRC) has been well-addressed by pattern matching, but the ability of common-sense reasoning remains a gap between humans and machines. Previous methods tackle this problem by enriching word representations via pretrained Knowledge Graph Embeddings (KGE). However, they make limited use of a large number of connections between nodes in Knowledge Graphs (KG), which can be pivotal cues for building the commonsense reasoning chains. In this paper, we propose a Plug-and-play module to IncorporatE Connection information for commonsEnse Reasoning (PIECER). Beyond enriching word representations with knowledge embeddings, PIECER constructs a joint query-passage graph to explicitly guide commonsense reasoning by the knowledge-oriented connections between words. Further, PIECER has high generalizability since it can be plugged into any MRC model. Experimental results on ReCoRD, a large-scale public MRC dataset requiring commonsense reasoning, show that PIECER introduces stable performance improvements for four representative base MRC models, especially in low-resource settings (The code is available at https://github.com/Hunter-DDM/piecer.).