A Comprehensive Study to Detect Social Spam Campaigns

Abstract

Social media is a widespread source of diverse forms of data. Information is diffusing so fast that any fascinating news is so rapidly spread through social media that it can be accessed by millions of users in seconds. Spammers are also part of these networks. Conventional spam detection approaches focus either on an individual message or individual account to classify them as spam or legitimate. However, these spam or spammers might be part of a group or controlled by some other accounts for a specific purpose called a campaign. Spam campaigns have become a great threat to social network services. They are more adversarial than individual spam accounts as they target many users on the network. Spam campaigns follow a stealthier approach than spammers to prevent themselves from detection. So, this study highlights various spam campaigns so far detected in the literature, their detection techniques, and the features used to build campaigns. It also highlights various issues and challenges that need to be picked up while detecting spammers.