Reference Matters: Benchmarking Factual Error Correction for Dialogue Summarization with Fine-grained Evaluation Framework

Mingqi Gao, Xiaojun Wan, Jia Su, Zhefeng Wang, Baoxing Huai