Optimization of E-commerce Product Recommendation Algorithm Based on User Behavior

Abstract

In order to implement personalized recommendation algorithms for e-commerce, the author proposes a genetic fuzzy algorithm based on user behavior to improve the sales, personalized recommendation, user satisfaction, and purchase matching performance of e-commerce. Collect data based on e-commerce personalized preference recommendation information, extract the associated feature quantities of personalized data for clustering processing, and then combine fuzzy B-means clustering method to achieve e-commerce personalized recommendation. According to the individual preferences of e-commerce, the collected data samples are fitted with differences and restructured, and a genetic evolution method is adopted for global optimization. The experimental results show that the optimized genetic fuzzy algorithm used in this method has improved stability and accuracy compared to the PSO method, with an  accuracy increase of 4%. This proves that the algorithm can provide the services needed by users more quickly and is an effective means.