The Design And Implementation Of Data Mining System Based On Logistics Information Platform

The rapid development of logistics information makes the scale of logistics data grown exponentially. Although the current logistics information platforms have a huge database, yet unable to analyze the data lacks effective analysis and mining, which makes the value and rules behind the data being hidden, so it is very necessary to build a data mining system based on the logistics information platform.In this paper, the design and implementation of data mining system is studied based on the national transportation logistics public information platform data. Firstly, the requirement analysis of the system is carried out and three major functional modules are designed: cluster analysis,transportation optimization and statistical analysis, finally the system is realized. It provides effective scientific basis for the decision of transportation departments, enterprises and individual users. The main work and achievements of this paper are:(1) Through sorting the existing data from the national transport logistics public information platform, the demand analysis based on the logistics information platform data mining system is carried out, the system construction goal is proposed. The paper has established the three function module of data mining: the module of dangerous goods transportation cluster analysis, the module of transportation optimization analysis and the module of statistical analysis. And it proposed more detailed demand scenes for every function module. A total of twelve data mining demand scenes are presented.(2) The data mining system based on logistics information platform is designed, including the overall design of the system, the design of system database and algorithm design. Among them, the system design part has designed the design scheme, the overall structure, functional modules and operational processes comprehensively. In the design of system database,the conceptual model of the database is described by using the E-R diagram,and by taking latitude and longitude data for grid division in our country,eight major data tables have been established ultimately. In the algorithm design part, it proposes an improved K-means algorithm for the clustering analysis of dangerous goods transportation from national logistics platform,the path optimization module combines the association rule algorithm and the A* algorithm to realize path optimization, the statistical analysis model utilizes Qliksense to analyze the goods industry and regional cargo capacity analysis.(3)The development realizes the data mining system based on logistics information platform, the system integrates three model functions,utilizes the data from logistics platform as data source for doing visual display of the results of eighteen mining and analysis scenarios, and compares the excavated knowledge with the results of traditional experience to verify The rationality of mining system.