BAG: Bi-directional Attention Entity Graph Convolutional Network for Multi-hop Reasoning Question Answering

Proceedings of NAACL-HLT 2019, pages 357–362 Minneapolis, Minnesota, June 2 - June 7, 2019. c⃝2019 Association for Computational Linguistics 357 BAG: Bi-directional Attention Entity Graph Convolutional Network for Multi-hop Reasoning Question Answering Yu Cao1 Meng Fang2 Dacheng Tao1 1UBTECH Sydney AI Center, School of Computer Science, FEIT, The University of Sydney 2Tencent Robotics X, China ycao8647@uni.sydney.edu.au, mfang@tencent.com, dacheng.tao@sydney.edu.au Abstract Multi-hop reasoning question answering re- quires deep comprehension of relationships between various documents and queries. We propose a Bi-directional Attention Entity Graph Convolutional Network (BAG), lever- aging relationships between nodes in an en- tity graph and attention information between a query and the entity graph, to solve this task. Graph convolutional networks are used to ob- tain a relation-aware representation of nodes for entity graphs built from documents with multi-level features. Bidirectional attention is then applied on graphs and queries to gener- ate a query-aware nodes representation, which will be used for the final prediction. Experi- mental evaluation shows BAG achieves state- of-the-art accuracy performance on the QAn- garoo WIKIHOP dataset. 1 Introduction Question Answering (QA) and Machine Com- prehension (MC) tasks have drawn significant attention during the past years. The proposal of large-scale single-document-based QA/MC datasets, such as SQuAD (Rajpurkar et al., 2016), CNN/Daily mail (Hermann et al., 2015), makes training available for end-to-end deep neural mod- els, such as BiDAF (Seo et al., 2016), DCN (Xiong et al., 2016) and SAN (Liu et al., 2017). How- ever, gaps still exist between these datasets and real-world applications. For example, reasoning is constrained to a single paragraph, or even part of it. Extended work was done to meet practical de- mand, such as DrQA (Chen et al., 2017) answering a SQuAD question based on the whole Wikipedia instead of single paragraph. Besides, latest large- scale datasets, e.g. TriviaQA (Joshi et al., 2017) and NarrativeQA (Koˇcisk`y et al., 2018), address this limitation by introducing multiple documents, ensuring reasoning cannot be done within local in- formation. Although those datasets are fairly chal- lenging, reasoning are within one document. In many scenarios, we need to comprehend the relationships of entities across documents before answering questions. Therefore, reading compre- hension tasks with multiple hops were proposed to make it available for machine to tackle such problems, e.g. QAngaroo task (Welbl et al., 2018). Each sample in QAngaroo contains multiple sup- porting documents, and the goal is selecting the correct answer from a set of candidates for a query. Most queries cannot be answered depending on a single document, and multi-step reasoning chains across documents are needed. Therefore, it is pos- sible that understanding a part of paragraphs loses effectiveness for multi-hop inference, which posts a huge challenge for previous models. Some base- line models, e.g. BiDAF (Seo et al., 2016) and FastQA (Weissenborn et al., 2017), which are pop- ular for single-document QA, suffer dramatical ac- curacy decline in this task. In this paper, we propose a new graph-based QA model, named Bi-directional Attention Entity Graph convolutional network (BAG). Documents are transformed into a graph …

Content truncated to 500 words