Application And Research Of Data Mining In The Customer Relationship Management System Of Correspondence Enterprise

The current correspondence market competition is intense, the user replaces the operation business to become a universal phenomenon frequently, does well the customer relations management to become the correspondence enterprise to maintain one of competitive power important attributes. However explodes facing the correspondence enterprise customer data -like growth, how is the correspondence enterprise enhances the data mining technology introduction customer relations management in the policy-making efficiency the key. The data mining can found the forecast customer behavior the model, helps the superintendent to extract the useful commercial information from the massive customer data, thus supports people's decision-making well. This article research goal is take Anshan correspondence company as a background, applies the data mining technology in the customer relations management, enables the customer management management system management system to support the enterprise decision-making well.First, this article the application elementary theory knowledge has carried on the comprehensive analysis to the data mining technology in the communication industry customer relations management, introduced separately the data mining and the customer relations management related concept, as well as the data mining in the customer relations management the application fundamental mode, discussed in CRM to carry on the main technical question which the effective data mining faced.Next, has constructed face the data mining correspondence enterprise customer relations management system management system overall structure and the realization frame, introducedthe subject analysis, the data storage management and so on each module function, had pointed out the data mining concrete application position, has analyzed the data mining process.Finally, based on Anshan correspondence company history in customer data, uses many kinds of data mining method to carry on the customer outflow forecast and the latent customer gain. In the customer outflow forecast, uses in the return forecast the multi-dimensional linear return establishment customer outflow forecast model, discovered and the outflow related customer attribute, obtains the customer expense custom and the outflow incidence relation. Uses the improvement the decision tree algorithm excavation forecast latent client base, and has given the gain latent customer reasonable feasible data mining flow. Through to forecast the result with the real time contrast, indicated the data mining model is effective.