Synthetic Data-Driven Approaches for Chinese Medical Abstract Sentence Classification: Computational Study
Our approach addresses the challenge of limited datasets for Chinese medical abstract classification by generating novel datasets. The deployment of SBERT-DocSCAN and SBERT-MEC models significantly enhances the precision of classifying Chinese medical abstracts, even when using synthetic datasets with pseudolabels.
Authors
- Longxuan Yu
- Z. Y. Wang (ORCID: https://orcid.org/0009-0001-5581-3207)
- Jiajia Li (ORCID: https://orcid.org/0000-0001-8158-3970)
- Haitao Song (ORCID: https://orcid.org/0000-0003-2113-2224)
- Hui Liu (ORCID: https://orcid.org/0009-0002-1404-746X)
Institutions
- Shanghai Jiao Tong University (CN)
- University of California San Diego (US)
- Beijing Academy of Artificial Intelligence (CN)
- Shanghai Civil Aviation College (CN)
Publication Details
- Journal
- JMIR Formative Research
- Published
- 2025-03-19
- DOI
- https://doi.org/10.2196/54803
- Citations
- 1
- Primary Topic
- Topic Modeling
- Type
- article
- Field-Weighted Citation Impact
- 2.48