A Customer Segmentation Research Based On Data Mining

In today's increasingly competitive market economy conditions, customers have become the most important resourced for enterprises. Understanding different customer groups' different preferences, shopping attitudes, price concept is the key point for successful marketing. Wise businessman and marketing officers will make marketing strategies effectively, according to the features of different groups in different cities and the life-cycle stages of different products. While enterprises subdivide their customers, usually they use the traditional statistical methods, but that is not conducive to the discovery of hidden knowledge in the data. To this serious problem, Data Mining provides a favorable technical support.Based on the telecommunications industry, this paper makes a preliminary analysis for Chinese telecommunication enterprises, the necessity and feasibility of customer segmentation of this industry. Against the features of this industry, a customer segmentation function model based on data mining has been designed. To the empirical analysis, the customer data of a telecommunications company in Guangdong province is selected as the subjects. After pre-data preprocessing, using K-means and Apriori algorithm to make cluster and association rules analysis. Though cluster analysis, customers are divided into four categories, then, make a second cluster subdivision for the large-share low-end customers. By association rules analysis, some consumer behavior rules of value-added business are found. Finally, in the light of the data mining results, some suggestions for marketing are given, which will play a very big significance for the company.