Model Based On The Loss Of Business Intelligence Telecom Customers

Nowadays, Global telecom market becomes more open and flexible. With the change of customer requirements, increase of operation complexity, of the techniques and transformation of market competition, telecom providers face the competition environment, which is more and more intensive and complicated."How to improve high quality service to attract and retain more customers, extend market share, reduce operation cost and increase income", have become the most important topic for the decision makers in telecom market. It forces the telecom corporations to find the great breakthrough in front of the challenges, either from the pressure of market competition, or the interior profit pressure of the enterprise. While Business Intelligence can help them realize the business object. Therefore, research in Business Intelligence has high market value in real applications.Based on Sichuan Telecom, this thesis studied the telecom project development, related theories of the business intelligence, such as data warehouse, OLAP analysis, data mining, and neural network, then we proposed a customer lost analysis model with Business Intelligence. This model can overcome the subjectivity of the traditional model which only use the key factors provided by technical staff. In this thesis, we first use these factors to build the data warehouse; then, we use the OLAP analysis to find the main factors which have more effect on customer lost and delete factors with minor effect; thirdly, we use the neural network to analyze every remaining factor, and find out the key factors with the 80/20 rule; finally, we use these factors as input, customer lost degree as output to construct neural network, and use data set to train the network until it gets convergent, which means the customer lost model has been constructed successfully.At last, we analyzed the model with the data set provided by Sichuan telecom. The analysis results showed that, the model proposed in the thesis is feasible, and can be used to guide customer lost analysis for Sichuan Telecom.