The Application Research Of Analytical Customer Relation Management System Based On Data Mining Method

Operational CRM is a kind of information management software, which combines three parts of business procedure sucn as selling, market and customer service. General Operational CRM only pays attention to combining resources inside enterprises and can't catch the turn of the market outside fast. Analytical CRM aims at increasing system's commercial analysis and aid decision ability. To realize the purport of Analytical CRM, an advanced data analytical method is needed to draw out useful information from a large number of trade data. Compared with traditional analytical method Data Ming method has the advantage of operating conveniently and the result easier to understand. The application research of analytical customer relationship management system Based on Data Mining Method can help enterprises solve the bottleneck of expanding profits to increase. So we researched on the most important teaches in constructing analytic CRM system: to obtain customer and to increase customer's value. Direct against the characteristic of the analyzing task and the requirements of enterprises, modeling algorithm is proposed. An Analytical CRM system is designed and implemented. The main work is listed here:(1) By extracting data from business database, a data warehouse facing the analyzing theme is designed and implemented. Extracting and washing the data in the business database of the Operational CRM, which is converted into the data warehouse. The data in the warehouse can be used as the source for the data mining process and can help to construct accurate useful models to predict.(2) In order to get new customers, the Application of Data Mining Method in obtaining potential customer is researched. An analyzing step to carry on obtaining potential customers is put forward. The procedure of structuring the model is designed. The model set up by the traditional decision tree algorithm is unstable and low in accuracy. By adopting boosing method, Adaboost C4.5 algorithm is designed and implemented based on traditional decision tree algorithm to satisfying enterprise's requisition for accuracy and stability of the model.(3) In order to increase customer's value, the Application of Data Mining Method in cross marketing is researched. An analyzing step to carry on cross marketing is put forward. The procedure of structuring cross marketing model is designed. All modes will be mined by the traditional sequential pattern mining method. By adopting Multi-dimensional attributes and other commonly used restraints, ERMSW algorithm is designed and implemented based on Prefixspan sequential pattern Ming algorithm toconvenient for analyst to control the modeling course and to dig out patterns analyst truly interested in. The improved algorithm can limit the condition to react the needs of analyst and can dig out the cross-marketing model, which analysts are really interested in.(4) Because of the implemented algorithms above, By adopting object-orient technique, an analytical CRM based on Data Mining method is designed and implemented.