The Management of E-Commerce Evaluation System Based on RBM and DBN Missing Rating Detection

Abstract

In e-commerce evaluation systems, missing evaluation data is a common problem. It can lead to fake reviews by malicious users, affecting users' decisions on products and services. Therefore, this study introduces Restricted Boltzmann Machines (RBM) and Deep Belief Networks (DBN) to fill in missing rating data for sparsely rated users. An iterative optimization ranking method is also used to improve user reputation values for identifying malicious users. The results show that on the Netflix dataset, the DBN model achieves an accuracy of 91.05% and an F1 score of 89.79%. On the Movielens dataset, the DBN model achieves an accuracy of 97.53% and an F1 score of 96.42%, which is a 13.08% and 12.73% decrease in accuracy and F1 score compared to the Support Vector Machine (SVM) model. On the Movielens-100 dataset, the DBN model achieves an accuracy of 86.11% and an F1 score of 84.27%, significantly outperforming the other two models. These results demonstrate the significant performance of the proposed method in data filling and malicious user detection in evaluation systems. It has important application value in the management of e-commerce evaluation systems.