Data Mining And Analysis Based On Industrial And Commercial Management Data Of Kunming

With the coming of the era of big data, the utility value of data is attract-ing increasing attention in various industries and businesses. Data mining, as an emerging technique, has shown remarkable advantages in data reuse, which can generate decent economic benefits to our life. To improve the management effi-ciency and adapt the modern management, the administrations for industry and commerce have gradually established their own affairs management systems, and achieved the digitization of economic accounts. However, the current management systems can only provide daily affairs management, which can not achieve in-depth data analysis and statistics as well as data mining since the explosive growth of data has greatly enlarged databases. Extracting valuable information from affairs management systems has become more and more difficult, while this has severely influenced the full utilization of data. Therefore, in order to deeply understand and analyze the industrial and commercial management (ICM) data, it is imperative to adopt data mining technique in the management systems.Based on the management data from the Kunming Municipal Administration for Industry and Commerce, and applying related techniques of data mining, this thesis demonstrated practical methods of data mining techniques for the ICM data, and presented the mining model of classifying the ICM data on the basis of classification analysis. Meanwhile, using the Java programming language and the three open-source frameworks including Spring, Struts and Hibernate as development tools, this thesis designed and realized a data mining system for the ICM data. Based on the above work, some application examples on the devel-opment trends of market entities were given. Moreover, according to the mining request, data characters and decision tree algorithm, an improved C4.5 algorithm is also proposed in this thesis. Compared with the unimproved algorithm, the new algorithm adopts attribute optimization and introduces user interest, which makes the mining results of the improved algorithm better meet the practical demands.