Kunming Telecom Fixed Telephone Customer Churn Characteristics Of Analysis

The paper mainly took the Kun Ming telecommunication business analysis supporting system as a background, took the telecommunication customer chum as an research object. The most serious churn is fix telephone, so we took it as a key point of research. According to needs, we completed the different granularity modeling for the data warehouse which based on the analysis of chum the subject, then, carried on to the data cube by the level multi-dimensional analysis and the deep level excavation. The paper took Kun Ming telecommunication two years chum data as research object, has used the corresponding Sample data in view of the different analysis stage, it is adopted the multiple perspectives analysis that the main characteristic of chum.Through to the product dimension Online Analytical Processing with exploring, Acquired the chum of the fix telephone (including residence telephone and office telephone) user that is the most serious at present. And adopting cluster method that the technology of data mining cluster analyzes the characteristic above analysis result. According to using decision tree to analyze the result of cluster with, obtained the characteristic which the partial customers churns.This article has used the data warehouse technology, the OLAP technology and the data mining technology churns to the fixed telephone carries on exploring analysis. Uses in the analysis from the surface to the spot, is thorough by the level, layer upon layer progressives principle. The practice proved, the article obtains the customer churns the characteristic to had the explanation, conforms to the objective reality, had the referenced value regarding the instruction enterprise precise marketing.