Research On Bank Customer Relationship Based On Complex Network

In recent years, commercial banks in China gradually developed into financial institutions with increasingly high profit. However, the competition among them is become fiercer and fiercer. Due to such intense competition in the market, banks have turned their goal from scale merit to client benefit and proposed a new management philosophy which is called “customer-focused”. As a result, the customers become the core resource of the banks. Therefore, how to manage the customer resource more accurately and efficiently and to analyze the relationship between the customers for the bank can maintain and develop the size of customer is one of the key problems which need to be solved by the banks in China.Customer relation analysis is a multidisciplinary research field of information technology and management science. Based on organizing and managing the customer's datum, banks provide corresponding service strategy by evaluating and analyzing the different customer's value to improve client satisfaction and loyalty. The traditional customer's techniques mostly assume that their clients are independent of each other and treat each of them as the basic research object. Moreover, this approach ignores the structural information among the client data. Facing with the characteristics of complexity and structural appeared in the information of customer transactions, it is difficult for the traditional analysis technology to satisfy the demand of the analysis of the massive transaction data since it is overly concerned about the individuals. Therefore, for the increasing scale of customers, we analysis the shortages of existed technologies and take the structure information among the customers into consideration, this thesis proposes a new method that combines the complex networks theory and technology with customer relationship analysis, which provides a new perspective and idea for banks to deal with the problem. The main contents of the research are as follows:(1) Firstly, this thesis elaborates the basic concept of customer relationship management, the complex networks theory and the analysis technologies.(2) Secondly, research on customer transaction data in the bank of a provincial capital is conducted from the perspective of complex networks. The thesis introduces and constructs the customer transaction network mode. Based on the definition of customer transaction relational network and the built model, we analysis the client trade network topology quantitatively, such as the node degree and degree distribution, average path length, cluster coefficient, and so on. The experimental results on real-world client trade network show that the client trade network has the characteristics of scale-free and small-world, and can help us to understand the client trade network more precisely.(3) After have verified the complexity of the trading network, the thesis identify the significant clients by using the complex network analysis technologies. In order to find the important clients and the core clients, we reorder the nodes according to their degree centrality and betweenness centrality.(4) Based on the experiment of the ranking of important clients in the trading network, the thesis simulates the important clients' loss by the way of the complex network invulnerability and calculates the selected criteria. The change of the structure indicator is analyzed in details.