Ontology Repository Construction Research Of Anti-terrorism Knowledge Graph Based On Open Source Data

To ensure the in-depth and continuous application of technologies such as machine learning and cloud computing,the Ministry of Public Security proposed the concept of "smart policing",which has constantly improved China’s Anti-terrorism intelligence level.However,open-source intelligence on Anti-terrorism comes from complex sources,highly heterogeneous and fails to effectively complement human intelligence.Besides,the construction of Antiterrorism big data has its drawbacks,such as difficult man-machine interaction and poor interpretability of machine learning.All these impede the "smart policing" construction and challenge the country’s Anti-terrorism intelligence,highlighting the need for knowledge engineering to enrich intelligence sources and serve big data construction.To practice the concept of "smart policing" and promote the successful launch of knowledge engineering,this paper probes into the construction of the core in Anti-terrorism knowledge graph,i.e.,Antiterrorism ontology repository,through literature analysis and comparative study.First,in the theoretical elaboration stage,it conducts a systematic review of relevant concepts and theoretical approaches,including the concept,architecture and value of Antiterrorism knowledge graph,as well as the composition,functions and categories of Antiterrorism ontology.After a comprehensive comparison of the strengths and weaknesses of different approaches,the seven-step method is adopted.Second,in the design stage,it follows the process of the modified seven-step method to identify the pain points of demand in Anti-terrorism intelligence,and carries out top-level design of the purpose,functions,data sources,requirements,and composition of the Antiterrorism knowledge graph ontology repository.Third,in the construction stage,with reference to corresponding design and industry standards,it introduces the seven-step method to build the Anti-terrorism knowledge graph ontology repository and refines it with Text Rank.Fourth,in the evaluation stage,which consists of four steps,it builds a Anti-terrorism dataset and validates representative ontologies with Deep KE;randomly selects test statements for inclusion testing;perform consistency checks;examines and refines the breadth,depth,and relevance of knowledge in the ontology repository by three quantitative metrics,i.e.,RR,AR,and LR.According to the results,the technical path can successfully construct the Antiterrorism knowledge graph ontology repository,which fully covers the knowledge types in Anti-terrorism intelligence and meets the Anti-terrorism intelligence requirements.Last,in the analysis and application stage,based on visualization analysis examples of the ontology repository in information retrieval and the construction limitations,improvements are proposed to identify technical and practical fits and put forward application strategies,including but not limited to the identification,interception and blocking of violent and terrorist audio and video with the help of IRT,the detection of terrorist financing chains through Dijkstra,the discovery of newly formed terrorist groups using OOKB entities,and the identification and inference of operational intentions in combination with matter graph.The intelligence construction of China’s Anti-terrorism intelligence is thus provided with application Antimeasures.Based on the demand of Anti-terrorism intelligence,this paper introduces the construction of knowledge graph ontology repository into Anti-terrorism,designs and explores the scientific construction path in light of the characteristics of knowledge in the field,which is expected to inspire the research methodology and construction of future Anti-terrorism intelligence.