Machine Learning Approaches for Fake Reviews Detection: A Systematic Literature Review

Abstract

These days, most people refer to user reviews to purchase an online product. Unfortunately, spammers exploit this situation by posting deceptive reviews and misleading consumers either to promote a product with poor quality or to demote a brand and damage its reputation. Among the solutions to this problem is human verification. Unfortunately, the real-time nature of fake reviews makes the task more difficult, especially on e-commerce platforms. The purpose of this study is to conduct a systematic literature review to analyze solutions put out by researchers who have worked on setting up an automatic and efficient framework to identify fake reviews, unsolved problems in the domain, and the future research direction. Our findings emphasize the importance of the use of certain features and provide researchers and practitioners with insights on proposed solutions and their limitations. Thus, the findings of the study reveals that most approaches focus on sentiment analysis, opinion mining and, in particular, machine learning (ML), which contributes to the development of more powerful models that can significantly solve the problem and thus enhance further the accuracy and efficiency of detecting fake reviews.