Analysis Of Mobile Group Customer Churn Based On Data Mining

With the coming of 3G and the reorganization of carriers, fierce competition leads to customer churn which will have a direct impact on KPI such as market share and profits. The famous "80/20" rule tells us that 80% of the sales comes from 20% of the clients. These are the high-value customers which we should be more concerned to retain.The status quo is that as the mobile industry are unique, such as the great amount of data, complexity of data structure, lack of customer information, as well as complexity of business and so on, the accuracy of the forecasts and their availability are not very acceptable. In order to improve the accuracy of the forecasts and their availability, it is necessary to further the research of the application of data mining technology in forecasting customer churn.The main tasks of the dissertation are listed as below:1. To build up the data mining mart for Prediction of VIP Customer Tendency through analyzing source data.2. To study comparatively the prediction methods of the Data Mining technology, carry out data mining and result analyzing by adopting C4.5 Decision Tree algorithm and thus realize the application of mining theme of the prediction of VIP customer in the mobile communication.3. To optimize the model through using the concept of "group customer purity ", offer advice on making applicable model for the prediction of customer tendency under the circumstances of large size customer groups and complicated business in the communication system and bring up discussions in criteria establishment and Decision Tree's comparably high dependence on the criteria.Through assessing and testing, this model has reached the demand for practical application.