Detecting Evangelists and Detractors on Twitter

Social networking websites provide a suitable environment for interaction and topic discussion. With the growing popularity of online communities, estimulated by the easiness with which content can be created and consumed, some of this content became strategical for companies interested in obtaining population feedback for products, personalities, etc. One of the most important of such websites is Twitter: recent statistics report 50 million of new tweets each day. However, processing this amount of data is very costly and a big part of it is simply not useful for strategic analysis. In this paper, we propose a new technique for ranking the most influential users in Twitter based on a combination of the user position in the network topology, the polarity of her opinions and the textual quality of her tweets. In addition, we develop and compare two methods for calculating the network influence. We also performed experiments with a real dataset containing one month of posts regarding soda brands. Our experimental evaluation shows that our approach can successfully identify some of the most influential users and that interactions between users are the best evidence to determine user influence.