Customer Segmentation Based On Neural Networks For Mobile Communication Industry In The 3G Era

With the increasing demand for mobile communication services, the competition among major operators becomes fierce. Because profit from traditional business is already saturated, the focus of mobile communications industry begins to shift from customer attraction to customer stock. The operators make efforts to carry out value-added business and try to make it a new growth point of profits. Only through detailed analysis and positioning of user groups and addressing their characteristics, can companies obtain a high return on investment. Nowadays, the domestic mobile communication business model is gradually shifting to an advanced paradigm which is customer-centric and data-centric, and information-based.The objective of current study is to target at huge customer resource of China Mobile in the 3G era and to designate target market based on customer needs and purchasing behavior. By integrating data from the operating, accounting, billing, customer service and other systems, unified data scales and data mining model for customer segmentation were designed according to customer segmentation theory. Subsequently, we conducted a data clustering using neural network algorithm to complete the research on customer segmentation for China Mobile. Data were collected from student customers in a certain university. After a series of data preprocessing of data cleaning, transforming and so forth, we obtained a reasonable breakdown of the final results through clustering analysis using Kohonen SOM neural network algorithm, and finally verified the segmentation model based on neural network.