Re-ranking Biomedical Literature for Precision Medicine with Pre-trained Neural Models
- Resource Type
- Conference
- Authors
- Li, Jiazhao; Murali, Adharsh; Mei, Qiaozhu; Vydiswaran, V.G. Vinod
- Source
- 2020 IEEE International Conference on Healthcare Informatics (ICHI) Healthcare Informatics (ICHI), 2020 IEEE International Conference on. :1-3 Nov, 2020
- Subject
- Bioengineering
Computing and Processing
Signal Processing and Analysis
Precision medicine
Biological system modeling
Conferences
Bit error rate
Medical services
Information retrieval
Informatics
biomedical retrieval
pretrained models
BERT
- Language
- ISSN
- 2575-2634
We propose a biomedical literature retrieval approach that incorporates a domain-specific BERT model as an auxiliary re-ranker. Experiments on TREC Precision Medicine dataset show its effectiveness in improving retrieval performance by 6.2% in inferred NDCG and 6.8% in R-precision over the best-published results. The contribution of this study is to provide evidence of incorporating BERT in a biomedical literature retrieval system, which serves the overall goal to improve the information retrieval for precision medicine.