Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information

Kyubyung Chae, Jaepill Choi, Yohan Jo, Taesup Kim


Abstract
A primary challenge in abstractive summarization is hallucination—the phenomenon where a model generates plausible text that is absent in the source text. We hypothesize that the domain (or topic) of the source text triggers the model to generate text that is highly probable in the domain, neglecting the details of the source text. To alleviate this model bias, we introduce a decoding strategy based on domain-conditional pointwise mutual information. This strategy adjusts the generation probability of each token by comparing it with the token’s marginal probability within the domain of the source text. According to evaluation on the XSUM dataset, our method demonstrates improvement in terms of faithfulness and source relevance.
Anthology ID:
2024.findings-naacl.117
Volume:
Findings of the Association for Computational Linguistics: NAACL 2024
Month:
June
Year:
2024
Address:
Mexico City, Mexico
Editors:
Kevin Duh, Helena Gomez, Steven Bethard
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1809–1820
Language:
URL:
https://aclanthology.org/2024.findings-naacl.117
DOI:
10.18653/v1/2024.findings-naacl.117
Bibkey:
Cite (ACL):
Kyubyung Chae, Jaepill Choi, Yohan Jo, and Taesup Kim. 2024. Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information. In Findings of the Association for Computational Linguistics: NAACL 2024, pages 1809–1820, Mexico City, Mexico. Association for Computational Linguistics.
Cite (Informal):
Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information (Chae et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-4/2024.findings-naacl.117.pdf