多源异构数据情境下公安知识图谱模型构建与智能应用

Construction and Intelligent Application of Public Security Knowledge Graph Model Based on Multi-source Heterogeneous Data

受传统数据管理模式以及查询方式的制约,现有的公安数据存储缺乏统一的标准规范,"信息孤岛"现象突出。因此,在公安案件关联、线索碰撞时,往往会发生数据找不到、衔接不上的窘境。多种异构数据情境下,公安知识图谱创建能够从多种数据源中异构提取知识,经过信息抽取、知识加工和知识融合,实现公安知识互联,使侦查有数据可查、有迹可循,加速侦查人员获取有效信息的效率,提高侦查人员对数据的掌控和利用水平。另外,利用公安知识图谱中的数据关系网,还可以与犯罪模型进行比对,挖掘出较高可信度的团伙或动态轨迹,提供给业务人员动态管控或预警分析研判。

Restricted by traditional data management models and query methods,the existing public security data storage lacks a unified standard and norms,and the phenomenon of"information islands"appears.When public security cases are connected and clues collide,data is often not found and cannot be connected.Under a variety of heterogeneous data scenarios,the creation of public security knowledge graphs can extract knowledge from multiple data sources heterogeneously,through information extraction,knowledge processing and knowledge fusion,realize public security knowledge interconnection,make investigations have data,and accelerate investigators’access to effective information The efficiency of investigators has improved the level of control and utilization of data by investigators.In addition,the data relationship network in the public security knowledge map can be compared with the crime model,and gangs or dynamic trajectories with higher credibility can be mined,and dynamic control or early warning analysis and judgment can be provided.