| 基于BERT的电子病历实体关系联合抽取研究 |
| 修订日期：2022-11-08 点此下载全文 |
| 引用本文：黄晓芳,陈剑秋,周祖宏,等.基于BERT的电子病历实体关系联合抽取研究[J].医学信息学杂志,2023,44(2):28-34 |
| 摘要点击次数: |
| 全文下载次数: |
|
| 基金项目:四川省科技厅重点研发项目“面向多语种的全球前沿技术智能汇集与应用平台”(项目编号:2021YFG0031)。 |
|
| 中文摘要:分析中文电子病历数据实体关系提取常用方法,提出一种基于双向编码器表征的实体关系联合抽取算法,使用级联解码器以及指针标注方法完成实体关系抽取及实体识别,实验结果证明该方法可有效抽取电子病历实体关系。 |
| 中文关键词:电子病历 关系抽取 联合抽取模型 自然语言处理 |
| |
| Study on Joint Extraction of Chinese Electronic Medical Record Entity Relationship Based on BERT |
|
|
| Abstract:The paper analyzes the common methods of entity relationship extraction for Chinese electronic medical record (EMR) data, proposes a joint extraction algorithm of entity relationship based on bidirectional encoder representation from transformers (BERT), which uses cascade decoder and pointer annotation method to complete entity relationship extraction and entity recognition. Experimental results show that the algorithm used in the paper can effectively extract the entity relationship of electronic medical record. |
| keywords:electronic medical record(EMR) relationship extraction joint extraction model natural language processing(NLP) |
| 查看全文 查看/发表评论 下载PDF阅读器 |
