Leveraging Large Language Models for Comprehensive and Interpretable Text Quality Evaluation through a Checklist-based Approach

CHECK-EVAL is a novel text evaluation framework that leverages large language models to generate and evaluate a checklist of key points derived from the source document and the candidate text, providing a structured and interpretable assessment of text quality.