融合相似度算法与预训练模型的中文电子病历实体映射方法研究  

Study on Chinese Electronic Medical Record Entity Mapping Method by Fusing Similarity Algorithms and Pre-trained Models

作　　者：冯凤翔 任慧玲[1] 李晓瑛[1] 王巍洁 王勖 张颖 FENG Fengxiang;REN Huiling;LI Xiaoying;WANG Weijie;WANG Xu;ZHANG Ying(Institute of Medical Information&Library,Chinese Academy of Medical Sciences&Peking Union Medical College,Beijing 100020,China)

机构地区：[1]中国医学科学院/北京协和医学院医学信息研究所/图书馆,北京100020

出　　处：《医学信息学杂志》2023年第5期45-50,共6页Journal of Medical Informatics

基　　金：科技创新2030——“新一代人工智能”重大专项课题“中文医学术语体系构建”(项目编号:2020AAA0104901)。

摘　　要：采用自标注中文电子病历标准数据集,融合相似度算法与预训练模型并分别应用于实体映射的候选实体生成和实体消歧阶段,对不同相似度算法和预训练模型的性能进行比较分析。提出基于别名间相似性改进药物类实体映射效果的方法,结合Jaccard相似度算法与BERT预训练模型,高效实现海量中文电子病历实体映射任务。The self-annotated Chinese electronic medical record(EMR)standard datasetisused,the similarity algorithms and pre-trained models are fused and applied to the candidate entity generation and entity disambiguation stages of entity mapping,and the performance of different similarity algorithms and pre-trained models is compared and analyzed.A method is proposed to improve the mapping effect of drug class entities based on alias similarity,and the Jaccard similarity algorithm and BERT pre-trained model are combined to efficiently realize the task of mapping the entities of massive Chinese EMRs.

关 键 词：实体映射 实体标准化 相似度算法 电子病历 BERT模型 

分 类 号：R-058[医药卫生]