Research On Loyal Customer Mining Based On Decision Tree Technology

Finding loyal customers is the first step that cultivating loyal customers and making them create lasting economy benefits for enterprise, decision tree provides an effective technology support for it. Paper researches loyal customer mining based on the related theories of customer loyalty index and decision tree technology, the real case of machine manufacturing enterprise is the background of it. Building customer satisfaction index system by analyzing the impact factors of customer satisfaction for A group, obtaining real data from company's existing customers by the way of questionnaire survey and pre-processing data. Selecting the valid data as training samples set and testing samples set, building the mining model of loyal customers by CART algorithm of decision tree technology and powerful Clementine mining tool of SPSS, and testing the accuracy of the model. Test result shows that this model has good accuracy, the prediction result is satisfactory and the model will help the enterprise discover loyal customers accurately and efficiently, and provide decision support for enterprise with fostering loyal customers and enhancing customers value.