Customer Segmentation For Online Retail Enterprise Based On Information Entropy

With the rapid development of e-commerce,as well as the promotion of“Internet +” strategy,B2 C business model is increasingly valued by the enterprises and consumers in China.Compared with the traditional retail enterprises,online retail enterprises have several advantages,such as the small proportion of fixed assets,large amount of potential customers,no limitation of space-time to product service and so on,which make customers the most important asset of online retailers.The effective implementation of CRM brings a new way of customer relationship management for online retail enterprises,and the key to implement CRM is to segment customers scientifically.In addition,the scientific and reasonable customer segmentation provides a basis for the implementation of differentiated marketing,precision marketing and so on,to enhance enterprises' competitive advantages effectively.Therefore,it is of great theoretical and practical significance to study the theory and method of customer segmentation.There are still deficiencies in customer segmentation index system and segmentation model in the existing study on customer segmentation of online retail enterprises.Sorting out related articles,find that the essence of customer segmentation in based on customer attributes and behavior characteristics.Thus,information entropy is introduced into the study of customer segmentation to build customer segmentation model,which combines customer segmentation index system in the basis of RFM model and information entropy.Finally,analyzing 541910 transaction data of one online retail enterprise to provide support for the theory and method of customer segmentation practices in online retail enterprises.Specifically,the main results of this study are as follow :(1)The objective weighting method entropy weight method is applied to determine the index weight of the RFM model,which solves the problem that theweight is too subjective.(2)The cross entropy is used to improve the K-means clustering algorithm.In traditional K-means clustering algorithm,the number of clusters and the cluster centers need to be determined in advance.However,through the calculation of entropy,this problem can be solved.(3)According to the characteristics of online retail enterprises,the writer combines K-means clustering algorithm which is based on cross entropy and improved RFM model establish online retail enterprise customer segmentation model.Using empirical research to analyze segmented customers from two aspects of customer values and consuming behaviors.Thus,through a more detailed result of customer segmentation to guide the business decision,we can provide decision support for precision marketing of the enterprises.This paper expands the application areas of information entropy,enriches the theory and method of customer segmentation,provides new tools and methods for customer segmentation of online retail enterprises,which offers certain references for other similar enterprises on customer segmentation.