A Robust Opinion Spam Detection Method Against Malicious Attackers in Social Media
This addresses the issue of spam detection for social media platforms and users, but it is incremental as it builds on existing graph-based methods.
The paper tackles the problem of spam attacks in online reviews, where malicious users attempt to deceive detection systems, and proposes a robust graph-based method that estimates honesty, trust, and reliability values, showing efficiency compared to other graph-based methods in case studies.
Online reviews are potent sources for industry owners and buyers, however opportunistic people may try to destruct or promote their desired product by publishing fake comments named spam opinion. So far, many models have been developed to detect spam opinions, but none have addressed the issue of spam attack. It is a way a smart spammer can deceive the system in a manner in which he can continue generating spams without the fear of being detected and blocked by the system. In this paper, the spam attacks are discussed. Moreover, a robust graph-based spam detection method is proposed. The method respectively estimates honesty, trust and reliability values of reviews, reviewers, and products considering possible deception scenarios. The paper also presents the efficiency of the proposed method as compared to other graph-based methods through some case studies.