| Now CRM (Customer Relationship Management) are being the hotspot of the world. It was derived from the model of customer-centric business. CRM is the way to maximize the benefit by managing the relationship between companies and customers. With the fast development of data capturing technologies, voluminous data has been accumulated in databases and is still increasing rapidly. However, the growing velocity is becoming even faster. In order to get interesting information, people have to plunge into the sea of data. At the same time our database systems have not provided us efficient tools as well as the capabilities in data input, update, basic analysis and display. The users have stayed in the stage in which they are drowning in data but starving for knowledge.Data mining or knowledge discovery in databases has been acting as a new research field which incorporates the thoughts of machine learning, statistical analysis, neural network, scientific visualization and data management. Digital neural network is an intelligent information technology which simulates the process of information in human brain. The most noticeable character of digital neural network is the self-organization and self-adaptation, with which the digital neural network can learn automatically and alter the algorithm with the change of data. Businesses can use the data to divide customers into segments based on some variables of current customer profitability. With the result businesses can enhance the loyalty of customers, make them satisfied and to buy more products.In this paper, the formal description of data mining and digital neural network was put forward, and the characteristics of these two kinds of knowledge were analyzed. It was also analyzed that the method of classifying the telecom customers and the detail customer-clustering algorithm based on the SOM described, and then the model was built which can classify the customers who use the broad band. At the end of this paper the author summarized the work and proposes the future research directions. |
