Personalized Recommendation Method for E-Commerce Platform Based on Data Mining Technology

In recent years, E-Commerce has been widely utilized and attracted more and more attention. Therefore, in this paper, we propose a novel collaborative filtering based personalized recommendation method for E-commerce platform. Firstly, we illustrate the proposed personalization recommendation system contains three modules: 1) Behavior record module, 2) Model analysis module, and 3) Recommendation algorithm module. In our proposed personalized recommendation algorithm, recommendation results are gained according to users rating score on various products. Moreover, users rating scores are represented as a user rating data matrix. Secondly, in our proposed recommendation algorithm, maximum margin matrix factorization is used to obtain personalized recommendation results by semi-definite programming solvers. Finally, experimental results demonstrate that our proposed algorithm can achieve higher accuracy than other methods, and can provide more suitable products for E-commerce users.