The Analysis Of High Data Traffic Telecom Customer Churn Prediction And Retaining Opportunity Assessment Based On Data Mining

With the advent of the mobile Internet era, competition between telecom operatorshave turned from ordinary communicate to data traffic service, high data traffic customersfor telecom operators become very important. With the competition in telecom market.How does the telecom operators retain these high-traffic customers in a effective way, hasbecome a hot issue concern to telecom operators. To retain customers effectively, it isnecessary to predict the probability of customers churn, analyzes the reasons of customerchurn and where does the customers go, and predict the probability of successful customerretention, Make effective retention strategies to reduce operational costs.Data mining techniques can use historical data, identify potential useful rules that existin the data. In this paper, data mining techniques was used to customer retention work.Based on the analysis of S City mobile customer data, according to the CRISP-DM datamining process, use Teradata Warehouse Miner data mining software to establish ahigh-traffic customer churn prediction model, predict customers’ churn possibilities.According to the statistical analysis of customers’ historical data, combined withexperience, summed up the prediction rules of customer churn reasons. Based on thecustomer churn reasons prediction rules, summed up the prediction rule of high-trafficcustomer churn whereabouts, predicted the reason and loss forecast whereabouts ofhigh-traffic churn customers.In this paper, based on the result of high-traffic customer churn model, according tothe high-traffic customer churn reasons and churn whereabouts, using factor analysis,established a model to evaluate high-traffic churn customer retain opportunity. Using ofthe high-traffic churn customer retain opportunity model’s results, combined with theexperience, we classify the high-traffic churn customer into for classes.They are not needto retain, more difficult to retain, retain and easy to retain. Provide a theoretical basis fortelecom operators to develop and maintain customer retention strategies.