CONSISTENT: Open-Ended Question Generation From News Articles

TL;DR

This paper introduces CONSISTENT, an end-to-end system for generating open-ended questions from news articles, addressing the challenge of producing questions that are answerable and faithful to the input text.

Contribution

The paper presents a novel system for open-ended question generation from news articles, along with an evaluation dataset and analysis of its effectiveness.

Findings

CONSISTENT outperforms baseline models in automatic and human evaluations.

The system generates answerable, faithful open-ended questions from news articles.

An expert-annotated dataset of open-ended questions is provided.

Abstract

Recent work on question generation has largely focused on factoid questions such as who, what, where, when about basic facts. Generating open-ended why, how, what, etc. questions that require long-form answers have proven more difficult. To facilitate the generation of open-ended questions, we propose CONSISTENT, a new end-to-end system for generating open-ended questions that are answerable from and faithful to the input text. Using news articles as a trustworthy foundation for experimentation, we demonstrate our model's strength over several baselines using both automatic and human=based evaluations. We contribute an evaluation dataset of expert-generated open-ended questions.We discuss potential downstream applications for news media organizations.