Exploiting abused trending topics to identify spam campaigns in Twitter

Twitter is an online social network (OSN) with approximately 650 million users. It has been fairly characterized as one of the most influential OSNs since it includes public figures, organizations, news media and official authorities. Twitter has an inherent simple philosophy with short messages, friendship relations, hashtags and support for media sharing such as photos and short videos. Popular hashtags that emerge from users’ activity are displayed prominently in the platform as Popular Trends. Unfortunately, the capabilities of the platform can be also abused and exploited for distributing illicit content or boosting false information, and the consequences of such actions can be really severe: one false tweet was enough for making the stock market crash for a short period of time in 2013. In this study, we make an experimental analysis on a large dataset containing 150 million tweets. We delve into the dynamics of the popular trends as well as other Twitter features in regard to deliberate misuse. We investigate traditional spam techniques as well as an obfuscated way of spam campaigns that exploit trending topics and hides malicious URLs within Google search result links. We implement a simple and lightweight classifier for indentifying spam users as well as spam tweets. Finally, we visualize these spam campaigns and investigate their inner properties.