An Approach of Personalized Recommendation for E-Commerce Websites Based on Sequential Patterns

摘要

More recently, Web usage mining has been proposed as an underlying approach for personalized recommendation.However, related work has focused attention on two techniques: clustering and association rule. This article proposes an approach of personalized recommendation for e-commerce websites based on sequential patterns. The approach utilizes Frequent (Contiguous) Sequences Graph to extract user interest view and generate recommendation set, which can effectively improve the precision of personalized recommendation.