FairytaleQA Dataset

FairytaleQA is a benchmark dataset of question-answer pairs derived from children's stories, designed to evaluate and improve both human and machine comprehension of narrative texts. Current research focuses on using FairytaleQA to develop and assess question generation (QG) models, exploring techniques like data augmentation and multi-question generation to enhance question diversity and quality, and investigating the presence of gender bias within the dataset and its impact on model performance. This work contributes to advancements in educational technology and natural language processing by providing a valuable resource for developing more effective and equitable reading comprehension tools for children.