Research on Online Detection Method of E-commerce False Comments Based on LBP

Abstract

With the rapid rise of e-commerce platforms, how to dig out false comments quickly and efficiently has become an urgent problem to be solved. This paper innovatively proposes a practical and effective online detection model for false reviews. Different from the previous offline detection methods, this model uses a sliding time window mechanism to realize the function of real-time monitoring and can dig out the relevant information of false comments in a short time. By extracting the review data in each time window, the measurement method of review similarity is defined, so that the relationship between reviews is quantified and the related review graph is established. Then the comment node is associated with random variables, and the comment graph is modeled as a Markov random field to generate the false comment online detection algorithm. The experimental results show that the accuracy rate and recall rate of the algorithm proposed in this paper have been improved.