Research on the Tendency of Consumer Online Shopping Based on Improved TOPSIS Method

Abstract

Analyzing and predicting the tendency of consumers online shopping is the precondition of providing personalized recommendation service, and has attracted more and more attentions. Most of e-commerce platform shave various types of products, and there exists tremendous difference in consumers’ occupation, education background and other personalized features. This Paper realizes a TOPSIS Method which is based on entropy and fuzzy numbers. Compared to the traditional TOPSIS method, with the Association Rules mining method of data mining, the improved TOPSIS solves the problem in traditional TOPSIS method which requires manual intervention during execution. In this study, to implement intelligent tendency predicting and analysis of consumers online shopping based on data driven, three steps is carried out. Firstly, the data mining method is leveraged to obtain the fuzzy weights of evaluation indicator through analyzing the electric business transaction data, and then a fuzzy decision-making matrix is established between product and consumer’s attribute; finally, a product category sequence which can indicate the tendency of consumer online shopping is established through calculating.