Identifying and Profiling Radical Reviewer Collectives in Digital Product Reviews

Abstract

Ecommerce sites are flooded with spam reviews and opinions. People are usually hired to impede or promote particular brands by writing extremely negative or positive reviews. It is usually performed in groups. Various studies have been conducted to identify and scan those spam groups. However, there is still a knowledge gap when it comes to detecting groups targeting a brand, instead of products only. In this study, we conducted a systematic review of recent studies related to detection of extremist reviewer groups. Most of the researchers have extracted these groups with a data mining approach over brand similarities so that users are clustered. This study is an attempt to detect spammers with various models tested by various reviewers. This study presents proven conceptual models and algorithms which have been presented in previous studies to compute the spamming level of extremist reviewers in ecommerce sites and online marketplace.