实体消歧综述  

Entity disambiguation:A review

作　　者：段宗涛[1] 李菲[1] 陈柘[1] DUAN Zong-tao;LI Fei;CHEN Zhe(School of Information Engineering,Chang’an University,Xi’an 710064,China)

机构地区：[1]长安大学信息工程学院,西安710064

出　　处：《控制与决策》2021年第5期1025-1039,共15页Control and Decision

基　　金：陕西省重点研发计划项目(2019ZDLGY17-08,2019ZDLGY03-09-01,2020ZDLGY09-02)。

摘　　要：实体消歧是将文本中出现的命名实体映射到一个已知的无歧义的结构化知识库中的技术.实体消歧是自然语言处理中的关键问题,对自然语言的发展起到重要作用.实体消歧对知识图谱构建、语义搜索、知识问答、推荐系统等应用有着重要的意义.对此,从实体消歧的定义、分类和相关研究基础出发,对实体消歧技术进行全面的解析.首先,对实体消歧的五元组定义进行说明,并给出实体消歧的常用分类以及相关研究基础;然后,分别对基于聚类的实体消歧、基于实体链接的实体消歧的研究内容以及研究现状进行详细综述;最后,对实体消歧的应用以及评测进行总结,并对未来研究方向进行了展望.Entity disambiguation is a technology that maps named entities that appear in text to a known unambiguous structured knowledge base. Entity disambiguation is a key issue in natural language processing and plays an important role in the development of natural language. Entity disambiguation is of great significance to the application of knowledge graph construction, semantic search, knowledge question answering, recommendation system and so on. Based on the definition, classification and related research basis of entity disambiguation, a comprehensive analysis of entity disambiguation technology is carried out. Firstly, the five-tuple definition of entity disambiguation is explained, and the common classification and related research foundation of entity disambiguation are given. Then, the research content of entity disambiguation based on clustering and entity disambiguation based on entity link, and the research status is reviewed in detail. Finally, the application and evaluation of entity disambiguation are summarized, and the future research directions are summarized.

关 键 词：知识库 知识图谱 实体消歧 自然语言处理 实体聚类 实体链接 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]