Medical Named Entity Recognition Based on Domain Knowledge and Posi-tion Encoding
Aiming at the difficulty of boundary identification caused by the large number of pro-fessional vocabulary in Chinese electronic medical record reports,this paper proposes a method to enhance entity boundary detection for better identifying medical named entities.The method takes entity boundary detection as an auxiliary task,so that the model can enhance the ability of entity boundary recognition,and then improve the effect of entity recognition.This paper en-hances entity boundaries from two perspectives.One is to introduce a self-made medical diction-ary to BERT for enhancing the ability to learn boundary information;the other is to use entity head and tail prediction as an auxiliary task to further enhance the model's ability to identify en-tity boundaries.Experiments are conducted on a public data set in the medical field.Comparing with the baseline model,the Fl value is improved by 1.96％,indicating that this method can ef-fectively detect entity boundary,improve the model performance,and verify the applicability of the model in the medical field.
Medical named entity recognitionnamed entity boundary detectionLEBERT
NETL
NSTL
万方数据
