STORYSUMM: Evaluating Faithfulness in Story Summarization
This work addresses the challenge of detecting subtle inconsistencies in narrative summarization for researchers in NLP, though it is incremental as it builds on existing evaluation methods.
The authors tackled the problem of evaluating faithfulness in story summarization by introducing STORYSUMM, a dataset with localized faithfulness labels and error explanations, and found that existing automatic metrics achieve less than 70% balanced accuracy on this task.
Human evaluation has been the gold standard for checking faithfulness in abstractive summarization. However, with a challenging source domain like narrative, multiple annotators can agree a summary is faithful, while missing details that are obvious errors only once pointed out. We therefore introduce a new dataset, STORYSUMM, comprising LLM summaries of short stories with localized faithfulness labels and error explanations. This benchmark is for evaluation methods, testing whether a given method can detect challenging inconsistencies. Using this dataset, we first show that any one human annotation protocol is likely to miss inconsistencies, and we advocate for pursuing a range of methods when establishing ground truth for a summarization dataset. We finally test recent automatic metrics and find that none of them achieve more than 70% balanced accuracy on this task, demonstrating that it is a challenging benchmark for future work in faithfulness evaluation.