The Application Of Data Mining In Customer Relation Management Of Huludao Unicom

At present,the competition in China's three major telecom operators is getting more and more intense along with the reformation of telecommunication market,so the customer relation management is becoming more and more crucial.According to the investigation a few days ago, the cost would be ten times higher to develop a new customer than to sustain a regular customer in telecommunications industry.Therefore,it would be a general trend to turn the customer relation management's center from business?setting complaints and aftermarket to avoiding the lost of customer.Then it's a big issue for operators to sustain existing customers rather than losing them.In order to provide the operator with a data support,by analyze their monthly spending,mining out the features of easy-to-lost customers from enormous amount customers,thus an efficient retain method would be used,and this is the significance of building a customer churn model.The traditional customer churn analysis mainly use single data mining,but the result is often one-sided,and have little guidance to practical application.So this thesis unite the recently customer relationship in communication company,constructs a customer losing forecasting T-N model based on decision-making tree and nueral network method.Basing on the customer data of China Unicom,Huludao Branch analysis on customers' losing.Thus bring out a high accuracy and efficient Unicomcustomer churn model.In addition,this thesis makes some feasible suggestion on customer relation management.