Research On Application Of Data Mining In MMS Operating Support

The application of data mining in MMS operating support is discussed in this paper. For the strategic target of improving the business development of MMS service, there are three key achievements in this paper:(1) MMS index prediction model based on BP neural network is proposed. The model can be used for forecasting the next phase of MMS indexes based on the holiday information, the number of current customer, the market gene and the historical indexes, and also can be used for market decision-making and instructing the development of MMS maintenance.(2) MMS customer classification model based on fuzzy clustering analysis which is used for customer classification is proposed. The model combines the information of customer's usage of MMS, customer profile and customer's usage of data service. Based on fuzzy clustering analysis, the model clusters customers into 4 categories: potential customer, high-activity customer, silent customer and maintainable customer. Enterprises could improve the customer value by precision marketing on high-activity customer and silent customer.(3)TSPA network quality upgrade model towards data service network quality customer sense is proposed. Aiming at customer recognition of data service network quality, the model analyzes the relationships between network quality management process, network quality technical indexes and customer recognition of data service network quality, and then establishes the expert knowledge base and rule sets for network quality.With these models above, the MMS operating supporting management platform is built. The platform provides the MMS maintenance with a standard management method and efficient support.