The Applieation Of Data Mining Technology In Telecom Customer Churn Alarm

Three domestic telecom operators increasingly competitive, provide a high degree of homogeneity of the product, leading to churn in the telecommunications industry is very high. How to effectively reduce customer churn is the operators need to solve the problem. By implementing IT systems to customers in the characterizations, through data mining to establish effective early warning models to improve customer churn churn warning mechanism to reduce wastage of users, carriers have become an important part of customer service. With the advent of mobile Internet, mobile Internet use continues to increase in population, the constant decrease in the use of traditional voice services, the user’s habits, behavior characteristics are changing. From the current situation in Guangxi certain operators, combined with the current widespread mobile Internet features, the online behavior data big data and traditional data combined with consumer behavior, build customer churn warning model, and through the assessment of customer churn prediction model and optimization to improve the accuracy of the model, for enterprises to reduce churn a useful exploration, providing effective IT support. This article focuses on the following aspects of the work done in.1. The status of the current operator churn analyzed,researched Customer churn data mining analysis.2. The main contents of the analysis of data to understand customer churn early warning and data preparation phase,Sources of research data, the Internet big data acquisition acquisition,Users online behavior data acquisition, analysis,Select churn warning model of the data field,Wide table design etc.The quality of the whole process of research data protection methods and propose solutions.3. Churn warning model was constructed by analyzing the evaluation and optimization models,By analyzing the evaluation and optimization model,To make effective churn warning retain work to maintain effective support.4. For Guangxi Branch conducted a pilot operator churn warning maintain retain work, the results of the work carried out to assessThis paper focuses on the theme of customer churn warning, will be applied to large data warning churn modeling, improved customer churn warning accuracy.May exist in data mining and data quality issues, the quality of the whole process of research data protection methods and propose solutions for data quality guarantee future data mining research has played a certain reference.