Fine-tuning pre-trained BERT-family models for candidate concept ranking improves biomedical entity normalization accuracy by up to 1.17% over prior systems on three datasets.
CNN-based ranking for biomedical entity normalization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
BERT-based Ranking for Biomedical Entity Normalization
Fine-tuning pre-trained BERT-family models for candidate concept ranking improves biomedical entity normalization accuracy by up to 1.17% over prior systems on three datasets.