Customer Classification Applied Research In E-commerce Environment

With the rapid development of science and technology, the use of the Internet and e-commerce has gradually penetrated into every area of socio-economic and cultural life. Human society continuely enter into an informational electronic and digital era. The change has brought new challenges to enterprise management, Customers become more and more mature and personality, as well as elusive. The relationship between enterprise and customers has been completely changed. The company’s core business of philosophy from "product-centric" gradually transformed into "customer-centric in the fierce competition. That is, only the first, fastest and the best service that can meet customers’needs and continue to attract new customers, keep old customers so as to remain invincible in the fierce competition. The emergence of customer relationship management achieves a good transition to "customer-centric" business philosophy. However, the most basic and core content in customer relationship management is customer classification. Using effective approach to classify the customers, identify customers accurately and differentiate different customer. Formulation and implementation of the different customer retention and customer upgrade the management and marketing solutions for different levels of customer groups,has important theoretical significance and application value. Based on the theory of data mining, management and marketing theory,in the e-commerce environment, customer classification applied research carried out a systematic theoretical research and case studiesThis thesis is focused on customer classification indicators, methods, reality, theoretical foundation and other related theories commence. A detailed study has been done. As well as an analysis of the problems of the customer applied research in e-commerce environment. Selecting log data of the network behavior, the customer value and outreach programs as indicators of the three dimensions of data, using D-SOM algorithm to construct the customer classification model in e-commerce environment. Utilization of K-means clustering algorithm again to classify on the same customer in order to verify the validity of the D-SOM algorithm, classification results tested by the Kruskal-Wallist that purpose to verify the results of the classification. The classification results are applied to customer retention to build customer loyalty fuzzy comprehensive evaluation model based on the coefficient of variation method; Customer classification results are applied to customer retention, using BP neural network algorithm to build a customer upgrade model, and ultimately get the best customer upgrade programFinally, as A search engine company for an example. Frist, it gets five kinds of VIP clients, training clients on focus, retention customer on focus, ordinary customers and small customers form50customers. At the same time, it analyzes each category of indicators trend in detail. Second, it evaluates customer loyalty of three kinds of clients as VIP customers, training clients on focus and key retention customer. At the end, this paper analyzes the reasons of low customer loyalty and help enterprises to find the direction of future efforts. The customer classification results are applied to customer retention. It chooses products, brands, prices, services and channels as five indicators, sets BP neural network algorithm to establish the model of the customer to upgrade, and ultimately get the best customer upgrade program for A search engine company.