A Novel Consumer Preference Model Based on Blockchain and Topic Similarity Clustering in Cross-Border E-Commerce

Abstract

Nowadays, countries in the world have frequent economic exchanges, and the scale of cross-border E-Commerce (CBEC) is getting larger and larger. CBEC refers to trading, payment, logistics, customs clearance, and other transactions. It also refers to the services between countries or regions through Internet technology and e-commerce platforms. In this article, the authors proposed a consumer preference model by blockchain and topic similarity clustering to obtain more consumer preference information. First, this article builds a blockchain-based consumer information collection system for CBEC to extract various features of consumers in CBEC. Secondly, by improving the performance of multiple characteristics of consumers, the accuracy of consumer preference prediction is improved. Finally, a method of consumer preference prediction based on topic similarity clustering is proposed to obtain consumers' purchase preference types. Experimental results show that the method can reach 84.3% of the H-mean, getting the best predictive performance and assisting CBEC by predicting consumer preferences.