CrossQG: Improving Difficulty-Controllable Question Generation through Consistency Enhancement

ACL ARR 2025 February Submission255 Authors

Abstract: Automatically generating questions with controlled difficulty has great application value, especially in the field of education. Although large language models have the capability to generate questions of various difficulty levels, the generated questions often fail to align with the given target difficulty. To mitigate this issue, we propose CrossQG, a novel training-free question generation method that enhances difficulty consistency. Specifically, CrossQG consists of two steps: (1) contrast enhancement, which leverages questions from different difficulty levels to enhance the base models' understanding of the target difficulty, and (2) cross filtering, which compares generated questions across different difficulty levels and filters out those that do not meet the target difficulty. We evaluate CrossQG on three high-quality question answering datasets, applying two difficulty estimation schemata. Experimental results demonstrate that across multiple models, CrossQG significantly outperforms several mainstream methods, achieving superior consistency with target difficulty and improving question quality. Moreover, CrossQG surpasses supervised fine-tuning in various instances even without training.

Paper Type: Long

Research Area: Question Answering

Research Area Keywords: question generation, reading comprehension

Languages Studied: English

Submission Number: 255