Buzzer Detection on Twitter Using Modified Eigenvector Centrality

Abstract

Social media is an online media where its users can easily participate, share, and create contents. One of the most used social media is twitter. Twitter nowadays used by billions of people to interact with other people. One of the phenomenon that we can observe in social media is user that has influence to other users, which commonly called influencer or buzzer. Buzzer often considered as central point of information spreading, which mean we can analyze it by using centrality analysis. Buzzer detection is one of problem that happen in social media that can be approach by using centrality analysis. One of the centrality analysis method is eigenvector centrality. Dynamics data that occur on twitter can be used as weight in eigenvector centrality and we made some modification in eigenvector centrality. On this paper, we propose a method by using modified eigenvector centrality to detect buzzer by considering dynamics data that occur on twitter.