Reference Matters: Benchmarking Factual Error Correction for Dialogue Summarization with Fine-grained Evaluation Framework
This work addresses the need for more reliable and detailed evaluation in factual error correction for dialogue summarization, though it is incremental as it builds on existing FEC methods.
The authors tackled the problem of unreliable evaluation for factual error correction in dialogue summarization by creating a manually annotated dataset of 4000 items and proposing FERRANTI, a fine-grained evaluation framework, which revealed significant performance differences across error categories and identified optimal training modes.
Factuality is important to dialogue summarization. Factual error correction (FEC) of model-generated summaries is one way to improve factuality. Current FEC evaluation that relies on factuality metrics is not reliable and detailed enough. To address this problem, we are the first to manually annotate a FEC dataset for dialogue summarization containing 4000 items and propose FERRANTI, a fine-grained evaluation framework based on reference correction that automatically evaluates the performance of FEC models on different error categories. Using this evaluation framework, we conduct sufficient experiments with FEC approaches under a variety of settings and find the best training modes and significant differences in the performance of the existing approaches on different factual error categories.