Research On The Construction Of Knowledge Graph Of Police Patrol

As a police basic business for the purpose of maintaining social stability and guaranteeing public safety,police patrol needs to obtain a large amount of information to assist the efficient and accurate execution of the business.The current utilization technology of police patrol data resources is relatively simple,which is difficult to support the associative use of multi-source heterogeneous data.Knowledge graph technology establishes the connection between data,forms a semantic network and locates more accurately the required information through semantic relations.The police patrol knowledge graph can be constructed to correlate the data related to the police patrol business field through semantic relations,and the multi-source heterogeneous data can be integrated and utilized.Based on the semantic network,the patrol can obtain rich and effective information and complete the patrol task quickly and accurately.This paper completes the construction of knowledge graph in police patrol,the main work is as follows:(1)Based on the ontology theory,constructing the ontology model of knowledge graph in police patrol.Through analysis of relevant data in the field of police patrol business,completing the classification of police patrol ontology and defining the attributes of the ontology and its value constraints,as well as the semantic relationship between ontology concepts.Data description of the model is realized by OWL(Ontology Web Language)language.(2)Knowledge acquisition of knowledge graph in police patrol.It mainly includes five aspects: data preprocessing,entity identification,relationship acquisition,knowledge fusion and knowledge storage.In the process of data preprocessing,the cleaning of network data,the transformation of semi-structured network data and the structuring of patrol-related laws and regulations data are completed.In the process of entity recognition,text data processing based on BILSTM-CRF model is completed,and entity selection based on ontology model is completed.In the process of relationship acquisition,completing the semantic relationship extraction based on Bootstrapping semi-supervised learning method and the relationship construction based on similarity calculation method.In the process of knowledge fusion,entity disagreement is eliminated;In the process of knowledge storage,data is transformed to RDF(Resource Description Framework)format and stored by D2 R.(3)The implementation and application of police patrol knowledge graph.The implementation of the ontology model is completed based on the DeepFinder and the knowledge graph is automatically generated through data import.Verifying the applicability of the graph through application examples in key areas and daily police patrols.The construction of the police patrol knowledge graph provides standardized,efficient and accurate guidance for the police patrol work,helps the training and guidance of the police patrol business,and improves the effectiveness of the police patrol business.