BertMCN: Mapping colloquial phrases to standard medical concepts using BERT and highway network

作者:

Highlights:

• Study the effectiveness of BERT based fine-tuned models to normalize medical concepts.

• Our best model based on Biomedical BERT and highway layer achieves state-of-the-art accuracy on two standard datasets.

• Study the impact of different learning rates, batch sizes and freezing encoder layers on our best performing model.

• Study the robustness of our best performing model against different noises.

摘要

•Study the effectiveness of BERT based fine-tuned models to normalize medical concepts.•Our best model based on Biomedical BERT and highway layer achieves state-of-the-art accuracy on two standard datasets.•Study the impact of different learning rates, batch sizes and freezing encoder layers on our best performing model.•Study the robustness of our best performing model against different noises.

论文关键词:Medical Concept Normalization,Clinical Natural Language Processing,BERT,Highway Network

论文评审过程:Received 21 November 2019, Revised 7 November 2020, Accepted 31 December 2020, Available online 7 January 2021, Version of Record 14 January 2021.

论文官网地址:https://doi.org/10.1016/j.artmed.2021.102008