3DM based customer classification algorithm of B2C

Any efficient customer relationship management is based on the excellent classification of customers.However,the correlated information of e-commerce customers is always huge,multi-dimension and incomplete,and it is hard to classify customers properly.The study shows that the essence of data mining is a knowledge transformation process in different formats.Thus,a data mining process is not only mining knowledge from data,but also from human.Domain-oriented data-driven data mining(3DM)can combine prior domain knowledge,user's interest and user's constraint with data sets effectively.According to the traits of e-commerce customers,a model of B2C customer classification was proposed in this paper.Furthermore,a new 3DM based algorithm was put forward,and the simulation results show its validity.