An Improved Apriori-Based Personal Recommendation Algorithm for E-commerce

Abstract

In the context of increasingly fierce business competition, the customers are overloaded with more and more product information. The e-commerce system is thus developed to provide customers with products of their interest, further maintaining customer loyalty, increasing brand sales and corporate competitiveness. Association rule-based recommendation is one of the most popular recommendation techniques for e-commerce. But it is challenging to extract association rules and the database is scanned unnecessarily for many times. To address these problems, an improved Apriori algorithm is proposed in this paper. An Apriori-based recommendation model is established. Results of MATLAB simulation demonstrate the ability of the proposed algorithm to provide customers with a list of products of their interests.