Research On E-commerce Customer Loyalty Based On Clustering Analysis

According to the characteristics of e-commerce industry, the influence factors of e-commerce customer loyalty are comprehensively analyzed, based on classical RFM model based on customer loyalty, establish customer loyalty of RFMSA e-commerce model. In this model, two important attributes affecting customer loyalty are introduced, which are the degree of satisfaction and the degree of attention. A more comprehensive analysis of the division of customer loyalty in electronic commerce transactions from multiple perspectives.On the basis of establishing the model of customer loyalty in electronic commerce, the customer loyalty is divided by the cluster analysis algorithm. Based on the classical clustering analysis algorithm K-means, the improved algorithm of the initial cluster centers is proposed to partition the customer loyalty. The method can effectively reduce the time of the initial cluster center and the local optimal solution generated by the iterative calculation.Using clustering algorithm to classify the customer loyalty, the existence of the classification boundary is not clear, in order to make more accurate classification of customer loyalty, this paper uses clustering algorithm based on rough set, and puts forward the rough clustering algorithm based on degree of membership. By using the concept of membership degree, not only the rough set and clustering method can be effectively combined, but also can make the point classification in the vicinity of the clustering boundary more reasonable. Through the analysis of the classic sample data, the experimental results show that the improved rough set K-means clustering algorithm can effectively improve the accuracy of clustering.In order to verify the effectiveness of the proposed algorithm, through the analysis of the transaction data of an online shopping mall, the customer loyalty is divided. Experimental results show that with K-means based clustering analysis algorithm compared to the accuracy of electronic commerce customer loyalty division of the improved algorithm is proposed in this paper is higher, can be better on the customer loyalty of e-commerce division.