RESEARCH ON INFORMATION SHARING AND FUSION OF PUBLIC SECURITY INFORMATION UNDER BIG DATA ENVIRONMENT

The public safety information system has built a large number of application systems, but there is a serious "information island" phenomenon. How to change this phenomenon has become an important issue. In order to realize the coordination and interaction between various application systems, we improve the sharing and integration of data and information in the public safety information system and connect various business processes and coordinate each police which lays the foundation for a systematic challenge and promotes the real practical value of big data in public safety information systems. Taking the acquisition and realization of data integration and sharing of fusion information resources as the principal purpose, we use data fusion theory to determine the knowledge integration of different dimensional levels of public safety information system website information resources and form a knowledge network according to specific topics of public safety information. The system website information resources are extracted and preprocessed with finegrained information, and the ontology-driven metadata model is used to carry out a unified semantic description. The integration and sharing of data knowledge in the field of public safety information system website information resources are established. Based on experiments under Linux environment, the LDA model in this paper collects information data on web page subject judgment and link relevance. Through experiments on 150 sample data, within the weight coefficient value range, the model is used for the subject text of public safety information system web pages. The data acquisition rate is as high as 91%. The same sample data is used for data fusion extraction of web keywords, and the N-Gram algorithm in the article has a keyword acquisition rate of 94.5%. The LDA model and the N-Gram algorithm collect 150 web pages from the public safety information system website and obtain the relevant web page theme harvest rates of the mentioned different topic collection algorithms. The research results show that the data information topic collection algorithm comprehensively utilizes the web topics. The dual filtering of judgment and link topic prediction can ensure that the relevance of the web page and the topic can be collected and the effective information and keywords in the web page can be effectively extracted. Therefore, compared with other algorithms, the web page topic harvest rate is higher overall.