Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Steps

TL;DR

This paper introduces 2WikiMultiHopQA, a new multi-hop QA dataset with explicit reasoning paths, designed to evaluate and improve models' multi-step reasoning abilities using structured and unstructured data.

Contribution

The paper presents a novel dataset with explicit reasoning paths, combining structured and unstructured data, and a pipeline ensuring high-quality, multi-hop questions for comprehensive evaluation.

Findings

The dataset is challenging for existing multi-hop models.

It provides explicit reasoning paths for better interpretability.

Questions are carefully designed to require multiple reasoning steps.

Abstract

A multi-hop question answering (QA) dataset aims to test reasoning and inference skills by requiring a model to read multiple paragraphs to answer a given question. However, current datasets do not provide a complete explanation for the reasoning process from the question to the answer. Further, previous studies revealed that many examples in existing multi-hop datasets do not require multi-hop reasoning to answer a question. In this study, we present a new multi-hop QA dataset, called 2WikiMultiHopQA, which uses structured and unstructured data. In our dataset, we introduce the evidence information containing a reasoning path for multi-hop questions. The evidence information has two benefits: (i) providing a comprehensive explanation for predictions and (ii) evaluating the reasoning skills of a model. We carefully design a pipeline and a set of templates when generating a question-answer pair that guarantees the multi-hop steps and the quality of the questions. We also exploit the structured format in Wikidata and use logical rules to create questions that are natural but still require multi-hop reasoning. Through experiments, we demonstrate that our dataset is challenging for multi-hop models and it ensures that multi-hop reasoning is required.