The Application Of Data Mining In Automatic Outbound System

With the consummation of restructuring works in China telecommunication industry, a framework of tripod stands is formed and the competition will be more and more intense in the communication market. How to maintain and keep inventory market, provide high quality and personalized service, and explore new market at the same time becomes competitive focus of telecommunication operators. Besides, the continual changes in user communication requirements and the continual falling of communication price result in the dropping of traditional voice income continually. Seeking for new business income growth poles is turn into an important task for each telecommunication operator.From the analysis of income, the traditional voice income is falling, at the same time, the value-added business income is increasing and the value-added business user-scale is extending. However, because of the abundant kinds and the fast development speed of value-added business in market, traditional business promotion and recommendation model can not meet the needs in actual situation. Under the above background, this article provides a new hidden marketing method - automatic outbound system, to make marketing to targeted customers about value-added business promotion and customized products service. In the selection of targeted customers, principal components analysis (PCA), cluster analysis and other data mining methods are used to get feature of targeted customer group, decision tree model is built for targeted customer, and finally regression testing is used to validate decision tree model. The testing result shows that the application of data mining in automatic outbound system effectively improves the phone answering rate and product ordering rate.The article firstly analyzes the current value-added promotion model. Then on the foundation of current model, a deep research on automatic outbound system, especially on data mining related technology, is made. The following works is completed in this article:1. Analyze the special requirements in value-added business promotion. Analyze the necessity of using data mining technology in building customer model for given business and selecting targeted customer in automatic outbound system.2. Introduce automatic outbound system in simple, and expatiate on problems on selecting the targeted customer. Introduce data mining technology and SAS data mining software.3. Emphatically introduce the framework design and work flow of automatic outbound system, and the design of data mining process.4. Introduce the hardware and software configuration of automatic outbound system in detail. It is verified by illustration that the automatic outbound system based on data mining has the ability to effectively improve the success rate of business ordering.