Rule Discovery Method And Its Application In The Customer Value Analysis

The current intense industry competition requires companies to convert their core running principle from focusing on product to focusing on customers. The customer relationship management(CRM) provides solutions to how to implement the running mode of focusing on customers. The core of CRM is the process management of mutual creation of the two types of value flow with different directions. One is from company to customer, the other is form customer to company. From the angle of the latter, the final aim of CRM is to maximize the customer value. It requires carrying out customer value analysis to maximize the customer value. This thesis applies data mining techniques to customer segmentation based on customer value matrix and builds the classification model of customer with different value.This thesis is unfolded according to the idea of discovering process of rules in the customer value analysis. Firstly, it introduces clustering concept and algorithm, compares the current popular algorithms and puts forward a modified algorithm of k-means, LKM. Secondly, it introduces the classification concept and algorithms and expatiates C4.5 with emphasis which is adopted in its customer value analysis model. Thirdly, it analyzes the connotation of customer value, brings forward the general process and core tache of customer value analysis and compares the current several methods of customer segmentation based on customer value. Lastly, it applies the innovated clustering algorithm and c4.5 to customer value analysis, brings forward a customer value analysis model based on customer value matrix, implements the customer value analysis of mobile company and makes experiments to compare the function of clustering algorithm and get the classification rules of customers of different value.