Research On Telecom Customer WLAN Service Order Prediction Based On Data Mining

While computer and database are widely and commonly used in every aspect of the society, huge amounts of data has accumulated. Although the amount of the data is vast, the knowledge is poor. The importance of data mining which canchange this situation has been recognized by data-intensive enterprise. Data Mining’s application and research will be wider.The application in telecom industry is a bit hot.In order to enlarge the scale of WLAN customer, and telecom operators need customer relationship management indeed.With the rapid development of the warehouse and data mining, the auther researched the prediction of WLAN business promotion activities in telecom industry based on data mining in the paper.This paper’s work is based on the project called’the sevice of Data Mining and Data Analysis’of K China Mobile Communications Corporation and the Data Mining techniques are applied to predict the WLAN business promotion activities. In the paper the author desiged and built the model of prediction of WLAN business promotion activities using customer data which got in the management analysis systems of Mobile Communications Company of K. This model is built under the frame of CRISP-DM, with the steps of business understanding, data understanging, data preparation, modeling evaluation and development. And SPSS Clementine is a tool which is used for Data Mining.First, the theory of Data Mining was introduced. And then effective data integration, reduction, cleaning, comversion and exploration were made before buliding the model to improve the quality of the data.Also, the author dealed with unbalanced data sets and initially established the model’s property index and data sets by the experience in business and attribute reduction.Finally, the model which used C5.0and CART of Decision Tree, Logistic Regression of two was built. And the author chose the optimal properties and the best model through training the model repeatedly and evaluation it with a number of indicators.During the modeling evaluation, the reason of business promotion activities was kown. And each customer could get a score from the model, then telecom operators promoted WLAN to the higher score customer. This is a good way to expand the scale of WLAN customer. At last the result indicates the forecasting model accords with the practical situation scientifically and can afford the predictive information and the solution project to decision-maker.This predictive model is importance increasing the number of WLAN customers.