The Design And Implementation Of Traceability System For Precursor Chemicals Based On Neo4j

At present,the Management Information System developed by Hebei Science and Technology Co.,Ltd.has accumulated a lot of management data for many years of uninterrupted service for public security management departments and related enterprises in many provinces and cities across the country.In recent years,the big data platform built by the company needs to be improved in many aspects,such as understanding enterprise relations,finding suspicious enterprises,tracking the source and destination of chemicals,such as analysis of the formation process of enterprise relations,accuracy and efficiency of product traceability,etc.The main work and innovation point of this thesis is as follows:1.Construction of enterprise relationship network.According to the actual needs,the data in the Management Information System is sorted out and updated.Combined with the characteristics of Neo4 j chart database,the enterprise relationship network is constructed with nodes and edges,which provides strong support for data analysis and management services in the future.2.Traceability optimization of precursor chemicals.In view of the problem that the original platform has not yet been able to quickly and accurately find the flow direction of all precursor chemicals.This thesis improves and optimizes the method of tracking the source and destination of precursor chemicals: make full use of the high efficiency of graph database query,adopt the idea of "cycle traversal,two-way output" to show the traceability results in the form of data visualization.The application effect of the optimized traceability method is good,and it is well received by users.3.Division of corporate community structure.Using GN algorithm,Newman fast algorithm and CNM algorithm,based on the characteristics of transaction network between enterprises,combined with the traceability results of precursor chemicals,the corporate community structure is divided.It provides the management with the functions of screening suspicious enterprises,dividing the key supervision objects and enterprise classification,and improves the supervision efficiency.Aiming at the problem of low accuracy of CNM algorithm,the similarity of nodes is used tooptimize CNM algorithm,which improves the accuracy of CNM algorithm.4.Research on the evolution of community structure.Adopt the idea that "the minimum time of precursor chemicals in and out of the warehouse is less than the deadline" and based on the enterprise relationship network of traceability,this thesis explores the evolution process of enterprise communities on the data sets at the provincial,municipal and district levels,which can not only obtain the strength of the enterprise's comprehensive strength and the type characteristics of closely connected enterprises,but also provide useful help for the enterprise to plan the direction of future operation and development,select partners,establish enterprise influence,etc.