Summary of Storysumm: Evaluating Faithfulness in Story Summarization, by Melanie Subbiah et al.
STORYSUMM: Evaluating Faithfulness in Story Summarization
by Melanie Subbiah, Faisal Ladhak, Akankshya Mishra, Griffin Adams, Lydia B. Chilton, Kathleen McKeown
First submitted to arxiv on: 9 Jul 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary In this research paper, the authors introduce a new dataset called STORYSUMM, which consists of summaries of short stories with localized faithfulness labels and error explanations. The goal is to create a benchmark for evaluating abstractive summarization methods, focusing on detecting challenging inconsistencies in summaries. The study shows that relying on a single human annotation protocol can lead to missing important errors, highlighting the need for diverse approaches. Furthermore, the authors test recent automatic metrics and find that none of them achieve more than 70% balanced accuracy on this faithfulness evaluation task. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The researchers created a new dataset called STORYSUMM that includes summaries of short stories with labels showing what’s correct or not in each summary. They want to use this dataset to see how well different methods can detect when a summary is not very accurate. The study shows that relying on just one way of looking at the data might miss some important mistakes, so they think it’s better to use multiple approaches. Additionally, they tested some automatic ways of measuring accuracy and found that none of them were able to get more than 70% right. |
Keywords
» Artificial intelligence » Summarization
