Abstract
Understanding the nature of high-quality summaries is crucial to further improve the performance of multi-document summarization. We propose an approach to characterize human-written summaries using partial information decomposition, which decomposes the mutual information provided by all source documents into union, redundancy, synergy, and unique information. Our empirical analysis on different MDS datasets shows that there is a direct dependency between the number of sources and their contribution to the summary.- Anthology ID:
- 2024.findings-acl.316
- Volume:
- Findings of the Association for Computational Linguistics ACL 2024
- Month:
- August
- Year:
- 2024
- Address:
- Bangkok, Thailand and virtual meeting
- Editors:
- Lun-Wei Ku, Andre Martins, Vivek Srikumar
- Venue:
- Findings
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 5333–5338
- Language:
- URL:
- https://aclanthology.org/2024.findings-acl.316
- DOI:
- Cite (ACL):
- Laura Mascarell, Yan LHomme, and Majed El Helou. 2024. Which Information Matters? Dissecting Human-written Multi-document Summaries with Partial Information Decomposition. In Findings of the Association for Computational Linguistics ACL 2024, pages 5333–5338, Bangkok, Thailand and virtual meeting. Association for Computational Linguistics.
- Cite (Informal):
- Which Information Matters? Dissecting Human-written Multi-document Summaries with Partial Information Decomposition (Mascarell et al., Findings 2024)
- PDF:
- https://preview.aclanthology.org/ingest-bitext-workshop/2024.findings-acl.316.pdf