Realtime social sensing of support rate for microblogging

Abstract

This paper proposes realtime estimation of support rate based on social sensors. Nowadays, micro blogs like Twitter have gained wide popularity, especially among the youth for its capability of updating personal opinions in a realtime manner. Academically, they have received tremendous attention as well. We argue that realtime events that have great influence on the attitudes of Twitter users can be detected by strategically monitoring tweets on certain topics. Building on the collected data, sentiment analysis enables us to calculate percentage of positive tweets, namely, support rate. In particular, given Twitter’s realtime nature, the support rate calculation shall also be done in realtime. Drawing on World Cup 2010, we collect a large amount of tweets and carry out analysis so as to extract sentiment information of the audience and go further to show the realtime support rate of the participators.