CRM Research Based On The Decision Tree Classification Algorithm

CRM is a kind of new management mechanism for aimed at improving the relationship between business and customer, It is mainly in the implementation of enterprise marketing ,sales ,services and Technical support and customer-related fields, make customers feeling business's exits all the time , business can see customers'change in time. Decision Tree Learning use since the top-down approach recursive , compare Internal node attribute value in the decision tree, then Judging from the node down branches by difference attribute value, given the result at the leaf node. So what from the root to leaf nodes of a path correspond a rule of admission, the entire decision tree correspond a disjunctive expression of a group of rules. By Decision Tree Algorithm can get the largest proceeds in enterprise Solutions in the base of haven data model.This study mainly researches that decision tree algorithm and the corresponding improved algorithm in CRM application. In the Data Mining Decision Tree Algorithm in the study ,the basic decision tree's algorithm has: CLS,ID3,CHAID,CART,FACT,SLIQ,SPRINT etc. , then Presents a more complete performance of the improved algorithm on the base of existing ID3 algorithm- can achieve optimal choice attributes, handling continuous in nature, without understanding the corresponding value of the ID3 algorithm attributes.At the process on CRM's Implementation this study are realized visual expression of the two algorithms under the classification of results, against ID3 algorithm and ID3 improvement. And under further CRM Data research's data collection in these two algorithms under the classification of the results of the two properties of the algorithm make analysis and comparison.