Multiple-Criteria Linear Programming for VIP E-Mail Behavior Analysis

Excessive lose of customer account is becoming a major headache for VIP E-Mail hosting companies. Analysis of features of easily lost customers and finding the appropriate measures to sustain them has become pressing needs for these companies. As a promising data mining approach, Multiple Criteria Linear Programming (MCLP) has been successfully applied to classify credit cardholders' behavior. The purpose of this paper is to apply MCLP model to VIP E-Mail behavior classification; we first extracted the features of VIP E-Mail according to the accumulated experience functions, then we tested MCLP with cross- validation and ensemble. The results of the study showed that MCLP is sufficiently accurate and stable on VIP E-Mail dataset. Therefore it can be concluded that the application of MCLP on VIP E-Mail behavior analysis can provide credible results.