Spammer Detection and Fake User Identification on Social Networks

Abstract

Millions of people from all over the world use social networking sites. The things that people do on social networking sites like Twitter and Facebook have a big effect on their daily lives, sometimes in a bad way. Spammers use popular social networking sites to spread a lot of useless and harmful information. For example, Twitter has become one of the most overused platforms of all time. Because of this, it lets a lot of spam through. Fake users send unwanted tweets to other users to advertise services or websites, which not only bothers real users but also slows down the use of resources. Also, fake identities have made it easier to spread false information to users, which makes it easier for harmful content to get out. Recently, finding spammers and fake users on Twitter has become a common topic of research in today's online social networks (OSNs). In this paper, we look at some of the ways that spammers on Twitter can be found. Also, a taxonomy of Twitter spam detection methods is given, which groups the methods by how well they can find: I fake content, (ii) spam based on URL, (iii) spam in trending topics, and (iv) fake users. The techniques that are shown are also compared based on different characteristics, such as user characteristics, content characteristics, graph characteristics, structure characteristics, and time characteristics. We hope that the study we are presenting will be a good way for researchers to find the most important new developments in Twitter spam detection in one place.