Data Mining Research And Application For Erp Data Mart

With the increasing popularity of the information technology, the ERP system has been widely applied to the process of manufacturing and management of the enterprise. However, the mass data accumulated during the running of the ERP system have not been made full use of. Data explode while the knowledge concerned is far from satisfying. With data mining, information critical to business decision-making can be elicited by transforming, analyzing and modeling the abundant database of the business concerned. Hence, in the course of enterprise modernization, data mining can enhance its competitiveness.This paper, by analyzing the present situation of the application of data mining, tries to study and develop ERP data mining on the basis of a purchasing decision supporting system of a company. The main work is as follows:Firstly, after the research on the general theories of data mining and the fundamental principles of ERP data mining, the paper gives a brief account of clustering analysis algorithm, RBF and neural network algorithm, and then focuses on the improved nearest neighbor clustering.Secondly, according to the characteristics of the management of raw material purchase, the author, by using the ERP system, presents the structure and function dividing of a purchasing decision supporting system, and then establishes a systematic information model based on dimensional modeling, and, by constructing an independent data mart, proposes ROLAP oriented toward purchasing decision.Thirdly, taking the need of making purchasing decisions of a manufacturing company into consideration, the author has developed a prototype purchasing system of ERP data mining by using PL/SQL and Delphi 7.0 in the Oracle 9i environment. The system with the function of supplier evaluation, has been put to actual use, the result of which is presented in the paper.Finally, the paper discusses and analyses the result, and presents the idea to improve the system.The result of the research has been put to use in the purchasing department of the company to help with its decision support, and has obtained desirable result.