Study On Behavior Mode Of Customers In CRM Of China Telecom

China telecom industry now are facing a series of changes caused from the business circumstances and the market composition. Under this new situation, they must pay attention to develop their core competence. CRM of customer as the center is suitable for China telecom industry. Only through CRM the companies know the customers' needs and supply the satisfactory products/service. Thus, the companies can exist and develop well. So, using CRM to rebuild their managerial mode is a necessity trend for china telecom industry to develop their information management. Data is the foundation of a good CRM. There is a great deal of important information concealed under the mass of data relevant to customers in corporation, which do well in the development of corporation. But at present time it cannot be found out speedily and available because of the disability of method being used by people. Data mining (DM) can set up the model which can forecast the trend of behavior of customers, help people to pick up the useful information from mass of data and thereby nicely support people's decision. Under this background, the use of the DM technology in CRM has important meaning. In this article we mainly research the technology of DM in CRM to analyze the behavior of customers. According to the information of customers and their phone bill in detail, recorded in the database of telecom enterprises, we classify them into different group. We study each group and try to find out the same characters of behaviors, such as capability, custom and the consumptive period of customers, in the same group and then form our rules of behavior mode of customers. Using those rules we can forecast the trend of behavior of customers and help people to make decision. The definition and the process of DM are introduced firstly in this paper, then the cluster and RBF NN are expatiated. We also expatiate the process of how to set up the Behavior Model of Customers using the cluster and RBF NN algorithm and the process of Intelligent Decision System. Finally we summarize our work and then point out the improvement of our work.