140 characters of @hate and #protest

Abstract

Research shows that cyber-hate, illegal or malicious form of cyber-protest and cyber-activism in online social media and Web 2.0 platforms has become a common phenomenon. This is a growing concern and hence automated techniques to counter such forms of online propaganda and identification of users and virtual communities is an area which has recently attracted a lot of research attention. In this paper, we present a simple and effective method to mine Twitter (a very popular and largest micro-blogging service) data for automatically identifying users and communities having a shared agenda. We propose a generic framework that consists of a systematic and focused traversal of the follower-network on Twitter and a user profile classifier based on contentbased features. We customize the proposed framework for a specific domain and demonstrate that the proposed approach is effective. We perform empirical analysis on data crawled from Twitter and the experimental results on the test dataset reveals that the proposed features and framework can successfully identify twitterers and hidden communities having a common agenda or shared interest.