| As the rapid development and extensive application of Data Warehouse and OLAT(On-Line Analysis Processing), Data Mining technology based on continuous increase of the amount of data and the decision requirement comes into being. It's a key component of Business Intelligence technology and it can serve to decision-makers according to the potential model found by corresponding algorithms. Facing the fierce competition in the market, Data Mining technology is destined to become one of the important methods to support company on database marketing service.This discourse discusses the massive database which coming from the HMJ(Huang Ma Ja) 96128 E-commerce platform of Chinese Business Newspaper, which sales goods by internet and call center. The problem of customer churn is solved by applying ID3 Algorithm, the problem of price is dealt with by applying K-means Algorithm, the research of cross-selling is solved by applying Apriori Algorithm.The data in HMJ-CRM's (Customer Relationship Management) database are used in our work. First after the data are cleared,100 thousand data in database are obtained. Next, the data are mined. ID3 algorithm and Apriori Algorithm are executed by Java programming language, and K-means Algorithm is finished by C++programming language. The results are obtained after running the programs and are represented by graphs.Experiment results show that paying attention to the analysis of customer churn, price of goods, shopping bag, the level of management would be raised radically and the business development of HMJ 96128 E-commerce platform would be stepped up. Ultimately it could service the leader scientific decision-making. |