Bots: Genuine or Malicious
Social network platforms are a major part of today’s life. They are used for entertainment, news, advertisements, and branding for businesses and individuals alike. However, the use of automated accounts also known as social bots pollutes this environment and hinders having a reliable clean online world. Malicious social bots manipulate public opinion. In this paper, we introduce genuine bots which are useful accounts that help the Twitter-sphere in some way. We call them genuine bots and examine their similarities and differences with malicious bots. We build a three-way classification between humans, malicious bots, and genuine bots on Twitter. Random Forest exhibits the best accuracy among all ML models. This work has great implications for social network analysis and the online sphere as using bots is increasing worldwide.
Committee: Jin Tian (major professor), Wallapak Tavanapong, and Adisak Sukul