Design And Implementation Of CRM Intelligent Customer System

Aimed at the case of the unfair assigned resource, combining Data Mining with CRM theory, a new CRM intelligent customer system was put forward applicable to A. A kind of technology of improved clustering and classification in Data Mining to mine association rules library of sale cost for underlying and useful knowledge was adopted in order to help manager schedule their service resource to meet their valuable clients. Hence, the modle solved the problem radically that the usful resource was caught in general client and valuable client can't get the sufficient service resource.There're some new ideas in this academic paper. Firstly, two kinds of CRM anlitical model was introduced in this paper which reflect the importance of client and reward rate compaird with those who concentrate on single target exactly to analyse the importance of client and sale cost. If you only consider the importance of client, maybe you will ignore some vital rules, which result in the situation of heavy and sightless investment. The ideas of sale cost can resolve the problem perfectly. On the one hand it can provide some detailed informations for enterprise, on the other hand it can offer some constructive suggestions on reward rate, whch resolved the problem radically that the usful resource was caught in general client and valuable client can't get the sufficient service resource.Except that, two kinds of improved classical algorithms were proposed on the basis of previous theory:the improved k-means algorithm based on square error density rule and the improved Apriori algorithm based on certainty frame. The improved k-means algorithm took the point with the biggest square error density as the initial center of mass, which solved the problem that the initial center of mass was chosen randomly or by some personal experience. Besides that, this paper also proposed a method which computed the quantities of K and solved the question that there's no method to know the quantities of K. when the algorithm begin to run. The improved Apriori algorithm put forward a new algorithm of association rules based on certainty frame, which adopted the certainty theory of MYCIN. This paper proved that the improved Apriori algorithm had better performance in some questions than traditional Apriori algorithm based on confidently frame through an example.