龙源期刊网 http://www.qikan.com.cn
基于知识图谱的人员关系预测方法研究
作者:朱振华 于晓昀 李超
来源:《电脑知识与技术》2018年第28期
摘要:针对大量数据表到知识图谱转化时,仅展示有直接关系的数据,无法挖掘关联有间接关系的数据的问题进行研究,基于公安业务场景,阐述了知识图谱搭建的流程,并以知识图谱关系预测的原理为基础,构建了符合公安行业的知识图谱关系信息资源模型,以规则预测+算法预测相结合的方式,提出了关系预测的技术路线,在海量数据中,实现实体的显性关系和隐性关系的深度挖掘和关系拓展。
关键词:知识图谱;公安业务;海量数据;关系预测
中图分类号:TP18 文献标识码:A 文章编号:1009-3044(2018)28-0176-03 Research on Prediction Method of Personnel Relationship Based on Knowledge Map ZHU Zhen-hua, YU Xiao-yun,LI Chao
(The First Research Institute of the Ministry of Public Security of P.R.C, Beijing 100048, China)
Abstract: In view of the problem that a large number of data tables can not be transformed into knowledge map only by displaying directly related data and mining indirectly related data, this paper expounds the process of building knowledge map based on the public security business
scenario, and constructs a knowledge map relationship prediction system based on the principle of knowledge map relationship prediction. The relational information resource model of knowledge map is presented. The technical route of relational prediction is put forward by combining rule prediction with algorithm prediction. In large amount of data, the deep mining and relational extension of explicit and implicit relations of entities are realized.
Key words: knowledge map; public security business; massive data; relationship prediction 1 背景
近年随着“人工智能”概念的再度活跃,除了“深度学习”这个炙手可热的名词以外,“知识图谱”无疑也是研究者、工业界心目中的又一颗“银弹”。简单地说,“知识图谱”就是以图形(Graph)的方式来展现“实体”、实体“属性”,以及实体之间的“关系”。知识图谱的主要目标是用来描述真实世界中存在的各种实体和概念,以及它们之间的关联关系。
相关推荐: