QA4QG: Using Question Answering to Constrain Multi-Hop Question Generation

TL;DR

This paper introduces QA4QG, a novel framework that enhances multi-hop question generation by integrating a question answering module with a pre-trained language model, significantly improving performance on the HotpotQA dataset.

Contribution

The paper proposes a QA-augmented BART-based framework for multi-hop question generation, leveraging QA constraints and pre-trained models to outperform existing methods.

Findings

QA4QG outperforms state-of-the-art models on HotpotQA

Increases BLEU-4 and ROUGE scores by 8 points

Demonstrates the effectiveness of QA constraints in question generation

Abstract

Multi-hop question generation (MQG) aims to generate complex questions which require reasoning over multiple pieces of information of the input passage. Most existing work on MQG has focused on exploring graph-based networks to equip the traditional Sequence-to-sequence framework with reasoning ability. However, these models do not take full advantage of the constraint between questions and answers. Furthermore, studies on multi-hop question answering (QA) suggest that Transformers can replace the graph structure for multi-hop reasoning. Therefore, in this work, we propose a novel framework, QA4QG, a QA-augmented BART-based framework for MQG. It augments the standard BART model with an additional multi-hop QA module to further constrain the generated question. Our results on the HotpotQA dataset show that QA4QG outperforms all state-of-the-art models, with an increase of 8 BLEU-4 and 8 ROUGE points compared to the best results previously reported. Our work suggests the advantage of introducing pre-trained language models and QA module for the MQG task.