电子病历自由文本实体关系抽取 |
投稿时间:2016-12-15 点此下载全文 |
引用本文:马敬东,梁力凡,夏晨曦.电子病历自由文本实体关系抽取[J].医学信息学杂志,2016,37(12):2-9 |
摘要点击次数: |
全文下载次数: |
|
基金项目:教育部网络时代的科技论文快速共享专项研究课题“基于互联数据的论文共享方法”(项目编号:0214516155)。 |
|
中文摘要:针对医学领域采用自举进行关系抽取的研究较少且国内面向医学领域的基础工具缺失问题,在一般自然语言处理技术的基础上,采用自举的算法框架,以最短依存路径构建关系模式,在过滤机制中引入候选实体的正向性评价,介绍新的算法优化策略,通过试验评价系统的性能,总结本研究的贡献与局限。 |
中文关键词:实体关系抽取 电子病历 自举 依存句法分析 |
|
Entity Relation Extraction for Free Text in Electronic Medical Records |
|
|
Abstract:Since there are few researches on relation extraction through bootstrap and few basic tools for medical field in China, on the basis of general natural language processing technology, the paper constructs a relation mode with the shortest dependency path and bootstrap algorithmic frame, introduces positive evaluation on candidate entities in the filtering mechanism and a new algorithm optimization strategy. It summarizes the contribution and limitation of the research by evaluating the performance of the system through tests. |
keywords:Entity relation extraction Electronic Medical Record(EMR) Bootstrapping Dependency parsing |
查看全文 查看/发表评论 下载PDF阅读器 |