Analysis Of The Telecom Market Network Complexity And Study Of Customer Segmentation Based On Social Network Analysis

Along with the deepening of reforming and restructuring, the operating marketenvironment of China Telecom has also undergone a fundamental change. The marketcompetition has become more and more intensified, especially in the recently years.Revenues which gained from expanding user number and decreasing price are not able tosupport the enterprises to survive and develop. During the process of seller marketchanging to buyer market, the enterprises have begun to take advantage of the intelligentdata analysis methods from different angles to analyze customers’ basic attributeinformation and consuming behavior information accumulated during daily operations, inorder to offer valuable information for scientific decisions. However, most of these methodsare based on customer’s individual attributes, ignoring the association among the customers.In recent years, along with the deepening research on the social network home and abroad,scholars discover statistical structure features of these social networks within complexnetworks, such as "small world effect","scale-free property" and so on. The telecom market,however, is a natural social network. It not only embodies the relationship between telecomoperators and customers, but also reflects the social relations between one people to anotherin some sense. Everyone in the real world is affected by the surrounding environment, soputting the customers in the social network environment has great practical significance.This paper uses social network analysis (SNA), and establishes the social relationshipnetwork of the telecom market with a telecom company’s real sales data in a year for threeconsecutive months as an example. Firstly, this paper analyzing the profile and structuralcharacteristics of the whole network using social network analysis method, which confirmsthe telecom market has the characteristics of "small world" and "scale-free feature". So thetelecom market network is a typical social network, and has the phenomenon of beingdivided into different groups. Then, based on the characteristic of the large number oftelecommunications customers, this paper improves the local community partitionalgorithm based on the number of common neighbor information. The improved algorithmonly needs to know the neighbors information of a node, without having to know theoverall features of the network. Therefore, it is suitable for customer segmentation of large-scale network. Finally, this paper applies the improved algorithm to subdivide China’stelecommunications market, and puts forward the corresponding marketing suggestions fordifferent segments.