High-Performance E-Commerce Personalized Recommendation System Based on Matrix Factorization and Convolutional Neural Networks

Abstract

Nowadays, how to effectively organize and use e-commerce information, maximize understanding of customer preferences, optimize website design, and make consumers more convenient is an important issue that urgently needs to be solved in the field of e-commerce. This article combined Matrix Factorization (MF) and Convolutional Neural Network (CNN) to construct an MF-CNN personalized recommendation model, and applied the MF-CNN model to a high-performance e-commerce personalized recommendation system. During the period from the 1st to the 11th, the overall click through volume of the website remained within a certain range. During the period from the 12th to the 22nd, there was a significant fluctuation in the click through of the popular product section on the homepage without using personalized recommendations. From the 23rd to the 29th, the website recommended products to users through personalized recommendations, and the click through of the popular product section increased to a certain extent. This high-performance e-commerce personalized recommendation system has a certain guiding effect on e-commerce recommendations.