Ask to Learn: A Study on Curiosity-driven Question Generation

TL;DR

This paper introduces Curiosity-driven Question Generation, a new task where questions are generated to seek new information not present in the input, using conversational QA data and novel evaluation methods.

Contribution

It defines a new question generation task focused on curiosity, proposes a methodology to create data from non-conversational datasets, and explores model approaches including pre-training and reinforcement learning.

Findings

Automated metrics to evaluate curiosity-driven questions

Models trained with pre-training and reinforcement learning improve question quality

Qualitative analysis shows relevance and inquisitiveness of generated questions

Abstract

We propose a novel text generation task, namely Curiosity-driven Question Generation. We start from the observation that the Question Generation task has traditionally been considered as the dual problem of Question Answering, hence tackling the problem of generating a question given the text that contains its answer. Such questions can be used to evaluate machine reading comprehension. However, in real life, and especially in conversational settings, humans tend to ask questions with the goal of enriching their knowledge and/or clarifying aspects of previously gathered information. We refer to these inquisitive questions as Curiosity-driven: these questions are generated with the goal of obtaining new information (the answer) which is not present in the input text. In this work, we experiment on this new task using a conversational Question Answering (QA) dataset; further, since the majority of QA dataset are not built in a conversational manner, we describe a methodology to derive data for this novel task from non-conversational QA data. We investigate several automated metrics to measure the different properties of Curious Questions, and experiment different approaches on the Curiosity-driven Question Generation task, including model pre-training and reinforcement learning. Finally, we report a qualitative evaluation of the generated outputs.