Algorithm Iid3 Crm Customer Classification And Research

Customer resources is the main point of the competition abilities to enterprises, more care of the sale groups (customer groups).take care of their opinion, requires and their propose of business, how to building a good, long term customer relationship improving the value of customers do good to the competition abilities and abilities to gain prices of the enterprises.In this paper, using the sale records of an enterprises as date sources, making a decision tree to classifying customers, getting the class rules, and mining out ideal customer. The base algorithm of decision tress is the ID3 algorithms. Its weak point is partial to the properties, which take more values. But these properties are not always the most valuable properties. In the course of calculating the entropy, the formula is complexes, the work of calculate is big, the complexion also higher. In this paper, during the course of using the ID3 algorithm to make decision trees, took out an improved ID3 algorithm. This algorithm get over the weak point of ID3 algorithm in some levels and getting an ideal model of decision tree classification.There also introducing the system design principle and implement scheme of customer relationship management, witch based on the B/S structures. Using the J2EE technology completes the system design.