Combing Customer Profiles for Members’ Repurchase Rate Predictions

Customer relationship management (CRM) leverages historical users’ behaviors in order to dawn effort of enhancing customer satisfaction and loyalty. Thus, constructing a successful customer profile plays a critical role in CRM. As customers’ preferences may change over time, we take the different types of past behavior patterns of the registered members to capture concept drifts. Then, we combine the repurchase index (RI) and the preference drifts to propose a Behavioral Repurchase Prediction (BRP) model, and to predict the members’ repurchase rates in the specific category of the e-shop. The marketers of the e-shop can target the registered members with high repurchase rates and design corresponding marketing strategies. The experimental results with a real dataset show that our model can effectively predict the registered members’ repurchase rates.