Answering while Summarizing: Multi-task Learning for Multi-hop QA with Evidence Extraction

TL;DR

This paper introduces the QFE model for multi-hop question answering that effectively extracts evidence sentences by considering their dependencies, improving explainability and achieving state-of-the-art results on HotpotQA and FEVER datasets.

Contribution

The paper proposes the Query Focused Extractor (QFE), a sequential evidence extraction model that captures dependencies among evidence sentences using an RNN with attention, enhancing multi-hop QA performance.

Findings

QFE achieves state-of-the-art evidence extraction scores on HotpotQA.

QFE also outperforms existing methods on FEVER dataset.

The multi-task learning approach improves explainability in QA systems.

Abstract

Question answering (QA) using textual sources for purposes such as reading comprehension (RC) has attracted much attention. This study focuses on the task of explainable multi-hop QA, which requires the system to return the answer with evidence sentences by reasoning and gathering disjoint pieces of the reference texts. It proposes the Query Focused Extractor (QFE) model for evidence extraction and uses multi-task learning with the QA model. QFE is inspired by extractive summarization models; compared with the existing method, which extracts each evidence sentence independently, it sequentially extracts evidence sentences by using an RNN with an attention mechanism on the question sentence. It enables QFE to consider the dependency among the evidence sentences and cover important information in the question sentence. Experimental results show that QFE with a simple RC baseline model achieves a state-of-the-art evidence extraction score on HotpotQA. Although designed for RC, it also achieves a state-of-the-art evidence extraction score on FEVER, which is a recognizing textual entailment task on a large textual database.