Research And Application Of Heterogeneous Data Integration And Clustering Mining

With the rapid development of information technology, the amount of information took explosive growth in every walk of life. It becomes increasingly difficult to acquire useful information and knowledge from substantial information resources which bear heterogeneous, massive and dynamic characteristics, and it brings great challenge in the fields of heterogeneous data mining. Therefore, we need data integration and data mining techniques to solve the problems in large-scale heterogeneous data mining for the purpose of eliminating Data Island and mining valuable information which become urgent problems in data mining system as well as a research focus in academia and industry.The paper applies the method of research-implementation-application to discuss the application of heterogeneous data source integration and clustering mining in the online games'heterogeneous customer data mining. Firstly, the paper points out the necessity and relevancy of the integration of heterogeneous data source and cluster mining. Secondly, the technologies of heterogeneous data sources integration and cluster mining are studied, and the existing typical techniques of them are introduced as well as their strengths and weaknesses are analyzed. On account of the study above, this paper proposes a heterogeneous data source integration technique based on grid service, the pivot of which is the design of the heterogeneous data source integration framework structure. And also this paper proposes a clustering mining algorithm based on density and grid, and it put emphasis on analyzing the quality and time efficiency of the algorithm proposed above, also conduct experimental data sets to verify the effectiveness and rationality of the algorithm simultaneously. Finally, the proposed heterogeneous data source integration technique and cluster mining algorithm are applied to an online game heterogeneous customer data mining of a company in ChongQing and a system of online game customer data mining is developed, which could dynamically integrate heterogeneous customer data of online games, subdivide customer data according to their consumption behavior and identify common characteristics, this provide the company for a precise marketing decision in the next step. The system results show that the integration of heterogeneous data sources and clustering mining technology which are used in the system of online game customer data mining are feasible and effective.In terms of the above research, the paper explores a practical-referenced new way for the application of heterogeneous data sources integration technology and clustering algorithm to the online games'heterogeneous customer data.