Select, Answer and Explain: Interpretable Multi-hop Reading Comprehension over Multiple Documents

Abstract

Interpretable multi-hop reading comprehension (RC) over multiple documents is a challenging problem because it demands reasoning over multiple information sources and explaining the answer prediction by providing supporting evidences. In this paper, we propose an effective and interpretable Select, Answer and Explain (SAE) system to solve the multi-document RC problem. Our system first filters out answer-unrelated documents and thus reduce the amount of distraction information. This is achieved by a document classifier trained with a novel pairwise learning-to-rank loss. The selected answer-related documents are then input to a model to jointly predict the answer and supporting sentences. The model is optimized with a multi-task learning objective on both token level for answer prediction and sentence level for supporting sentences prediction, together with an attention-based interaction between these two tasks. Evaluated on HotpotQA, a challenging multi-hop RC data set, the proposed SAE system achieves top competitive performance in distractor setting compared to other existing systems on the leaderboard.

Accepted to AAAI 2020

References in corpus (5)

• Dynamic Coattention Networks For Question Answering
• Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering
• Densely Connected Attention Propagation for Reading Comprehension
• Understanding Dataset Design Choices for Multi-hop Reasoning
• Evidence Sentence Extraction for Machine Reading Comprehension

Cited by in corpus (7)

• Longformer: The Long-Document Transformer
• Multi-hop Question Answering via Reasoning Chains
• Machine Reading Comprehension: The Role of Contextualized Language Models and Beyond
• Is Graph Structure Necessary for Multi-hop Question Answering?
• A Survey on Explainability in Machine Reading Comprehension
• Graph Sequential Network for Reasoning over Sequences
• Machine Reading Comprehension using Case-based Reasoning