Journal article
Word-Sequence Entropy: Towards uncertainty estimation in free-form medical question answering applications and beyond
Engineering applications of artificial intelligence, v 139, 109553
Jan 2025
Abstract
Uncertainty estimation is crucial for the reliability of safety-critical human and artificial intelligence (AI) interaction systems, particularly in the domain of healthcare engineering. However, a robust and general uncertainty measure for free-form answers has not been well-established in open-ended medical question-answering (QA) tasks, where generative inequality introduces a large number of irrelevant words and sequences within the generated set for uncertainty quantification (UQ), which can lead to biases. This paper proposes Word-Sequence Entropy (WSE), which calibrates uncertainty at both the word and sequence levels based on semantic relevance, highlighting keywords and enlarging the generative probability of trustworthy responses when performing UQ. We compare WSE with six baseline methods on five free-form medical QA datasets, utilizing seven popular large language models (LLMs), and demonstrate that WSE exhibits superior performance in accurate UQ under two standard criteria for correctness evaluation. Additionally, in terms of the potential for real-world medical QA applications, we achieve a significant enhancement (e.g., a 6.36% improvement in model accuracy on the COVID-QA dataset) in the performance of LLMs when employing responses with lower uncertainty that are identified by WSE as final answers, without requiring additional task-specific fine-tuning or architectural modifications.
[Display omitted]
•We devise Word-Sequence Entropy for uncertainty analysis in medical query-answering.•We investigate the issue of generative inequality in medical responses.•We capture and highlight keywords and reliable sequences based on semantic relevance.•We resample based on the calibrated uncertainty.
Metrics
Details
- Title
- Word-Sequence Entropy: Towards uncertainty estimation in free-form medical question answering applications and beyond
- Creators
- Zhiyuan Wang - University of Electronic Science and Technology of ChinaJinhao Duan - Drexel UniversityChenxi Yuan - University of PennsylvaniaQingyu Chen - University of New HavenTianlong Chen - Massachusetts Institute of TechnologyYue Zhang - Drexel UniversityRen Wang - Illinois Institute of TechnologyXiaoshuang Shi - University of Electronic Science and Technology of ChinaKaidi Xu - Drexel University
- Publication Details
- Engineering applications of artificial intelligence, v 139, 109553
- Publisher
- Elsevier
- Number of pages
- 12
- Grant note
- National Key Research & Development Program of China: 2022YFA1004100
Zhiyuan Wang and Xiaoshuang Shi were supported by the National Key Research & Development Program of China under Grant (No. 2022YFA1004100) .
- Resource Type
- Journal article
- Language
- English
- Academic Unit
- Computer Science
- Web of Science ID
- WOS:001359282700001
- Scopus ID
- 2-s2.0-85208764309
- Other Identifier
- 991021961001904721
InCites Highlights
Data related to this publication, from InCites Benchmarking & Analytics tool:
- Collaboration types
- Domestic collaboration
- International collaboration
- Web of Science research areas
- Automation & Control Systems
- Computer Science, Artificial Intelligence
- Engineering, Electrical & Electronic
- Engineering, Multidisciplinary