• Data Augmentation for Few-Shot...

*Result*: Data Augmentation for Few-Shot Biomedical NER Using ChatGPT.

Title:
Data Augmentation for Few-Shot Biomedical NER Using ChatGPT.
Authors:
Mu W; Dalian Minzu University, Dalian, 116650, China., Zhao D; Dalian Minzu University, Dalian, 116650, China. Electronic address: zhaodi@dlnu.edu.cn., Meng J; Dalian Minzu University, Dalian, 116650, China., Chen P; Dalian University of Technology, Dalian, 116024, China., Sun S; Dalian Minzu University, Dalian, 116650, China., Yang Y; Dalian University of Technology, Dalian, 116024, China., Wang J; Dalian University of Technology, Dalian, 116024, China., Lin H; Dalian University of Technology, Dalian, 116024, China.
Source:
Artificial intelligence in medicine [Artif Intell Med] 2026 Feb; Vol. 172, pp. 103314. Date of Electronic Publication: 2025 Nov 29.
Publication Type:
Journal Article
Language:
English
Journal Info:
Publisher: Elsevier Science Publishing Country of Publication: Netherlands NLM ID: 8915031 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1873-2860 (Electronic) Linking ISSN: 09333657 NLM ISO Abbreviation: Artif Intell Med Subsets: MEDLINE
Imprint Name(s):
Publication: Amsterdam : Elsevier Science Publishing
Original Publication: Tecklenburg, Federal Republic of Germany : Burgverlag, c1989-
MeSH Terms:
Data Mining*/methods , Natural Language Processing*, Humans ; Machine Learning ; Databases, Factual ; Algorithms ; Generative Artificial Intelligence
Contributed Indexing:
Keywords: Biomedical named entity recognition; ChatGPT; Data augmentation; Few-shot learning; Natural language processing; Prompt learning; Separable convolutions; Transfer learning
Entry Date(s):
Date Created: 20251203 Date Completed: 20260101 Latest Revision: 20260101
Update Code:
20260130
DOI:
10.1016/j.artmed.2025.103314
PMID:
41338038
Database:
MEDLINE

*Further Information*

*Data Augmentation (DA) aims to create a new dataset to address the lack of data in various domains. Particularly in few-shot scenarios of the biomedical Named Entity Recognition (NER) domain, an effective DA method can enhance data diversity, reduce overfitting, and significantly improve the model's generalization ability. In this work, we propose a novel DA method for NER tasks, which uses ChatGPT and prompt learning to extract high-quality data from large language models. The entity recognition tasks are then performed via transfer learning and efficient decoding strategies. Moreover, this study conducted extensive experiments on four publicly available biomedical datasets (BC5CDR, NCBI, BioNLP11EPI, and BioNLP13GE), demonstrating that our methods exhibit strong stability and entity recognition capabilities even in extremely limited scenarios. In the 5-shot, 20-shot, and 50-shot scenarios, the average F1 scores of the four datasets reached 72.96%, 75.05%, and 77.42%, respectively.
(Copyright © 2025 Elsevier B.V. All rights reserved.)*

*Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.*