Abstract: In order to solve the problems of automotive customers behavior prediction to predict the marketing trends accurately,we adopted 16 customerspurchase information as our sample data, and cleaned up and summed the data based on the theory of CRM(Customer Relationship Management), using the SQL Server we built the automotive marketing data warehouse model as the input of the whole process. We adopted the improved ID3 decision tree model and the association rule model to do data mining, and got information of market-predicting. And verified the validity of our prediction results through experiments and contrasts.

Key words: prediction of automobile customer behaviors, decision tree model, association rule model, data mining, customer relationship management (CRM)