Study And Application Of Demand Forecast Model In CRM Based On Data Mining

CRM is a hot spot of present business administration, in order to improve customer satisfaction, and then promote the level of profits, enterprises must grasp perfect customer's information and hold accurate customer demand , then make one's own products attractive, make one's own enterprise competitive purpose to achieve finally. Should realize this goal, it need imply useful information by drawing out from a large amount of data, need advanced the method of data analyzing and the method of demand forecast, and the technology of data mining has more advantages than the traditional method because it is convenient and easy to understand, the application study based on demand forecast model of CRM can help the enterprises to solve the bottleneck problem of profits and increase , so it is paid attention to more and more by researcher.Because the model of customer demand forecast need to be improved urgently, this paper has carried on research to describe customer and predict customer demand in which the most valuable link how to structuring CRM, this subject regards studying in probability as the theoretical foundation, according to domestic enterprise's situation and forecast demand at present, adopt the modeling method of mathematics to propose the new forecast model, which is divided into customer demand forecast model and product competition analysis model. Meanwhile, needed on the basis of the data of forecast model, chose the suitable algorithm after analyzing the advantages and disadvantages of the typical clustering algorithm synthetically, and has improved CHAMELEON algorithm by the methods of object-oriented and concurrent designing. Prove through the experiment, these models and improving have very strong practicability and efficiency, can apply it to various kinds of research work and practice which including comparison or competition. In the end, we design and realize a demonstration application system to prove and show the systematic function of the design, in order to promotion and application.