Research For Vendor Selection Decision Support System Based On Data Mining

Since the nineties of twenty century, with the rapid development of the information technologyand productivity, the global market competitions become increasingly fierce, customer’s demandsbecome more and more personalized, the importance of the supplier evaluation and selectionbecomes more and more outing in the supply chain management. For an enterprise, how to makethe mutual cooperation, win-win situation when selecting the supplier is one of the most importantproblem. In this thesis, the supplier evaluation indicator system, algorithm design and systemimplementation in medium and small enterprises are constructed. The main research is included asfollowing：(1) The construction of supplier evaluation index system. Combining the researchachievements on supplier evaluation from both home and abroad，This thesis established supplierevaluation index system and set up qualitative and quantitative analysis for all indexes in thesupplier evaluation index system.(2) The research of supplier selection methods based on data mining. Through analyzingexisting supplier selection methods, the process and methods of supplier selection based on datamining are determined. First of all, suppliers are classified by the FCM clustering algorithm. Thenthe weights of various indexes are determined to calculate the group of the optimal supplier throughAHP. Finally the optimal supplier is calculated and the empirical analysis is set up.(3) Design, development and analysis of vendor selection system. Based on the old ERPsystem developed by the development group including the author, B/S structure and JAVA language(ExtJs+Struts2+Hibernate+Spring) are adopted to develop the system. Comparing to the old system,the new one has expanded in the following two modules: basis data management of supplier andvendor selection.