Research On B2C Enterprise Customer Relationship Management Based On The Technology Of Data Mining

With the rapid development of e-commerce and information technology, data mining technology has been widely used in the business setting. And it can improve ability of online merchants to identify, analyze customers’ data and meet their requirements. But due to the differences of data collections, data analysis processing and outcome feedback in diversified fields, unique models and analysis method must be adopted in different industries and even enterprises.The purpose of this article is to improve the customer relationship management by means of taking the comprehensive use of several data mining technology in the customer relationship management (CRM) and it is feasibility on the basis of analyzing the B2C customers’consumption characteristics in e-commerce theory.The article analyzes the characteristics of B2C customers’consumption habits based on introducing the related concepts and technological means firstly, emphatically analyzes the process of customers’ analysis, and respectively analyzes the all customers’data analysis process in RFM and classifies the customer category with FRM on the basis of the customers’ attributes, the customers’ psychological characteristics and their network effects and other valueless index. Then it builds a customer classification model based on two-dimensional clustering, which is important to analyze the potential needs of customers to maintain attraction based on different classification. The article focuses on making use of association rules mining and analysis, and tentatively uses Apriori algorithm of association rules aimed to reduce number of scanning. For new customers, according to the basic registration information and browsing history it uses personalized recommendation technologies and attempts to consider the effect of the amount of data into scale factor to improve the calculation of correlation coefficient and enhance the accuracy of prediction. Thirdly, it constructs out 5 kinds of customer relationship management strategy using association rules of classification methods, respectively they are bolter, primary, ordinary, potential and golden, and it proposes the implementation of protection. At last, case analysis proves the effectiveness of proposed methods.This article attempts to adopt the way of double clustering to combine the classification of the customers, and explore the algorithm of improved association rule and recommendation system to make certain support to improve B2C enterprise customer relationship management level.