Modelling political disaffection from Twitter data
Proceedings of the Second International Workshop on Issues of Sentiment Discovery and Opinion Mining
| Hi-index | 0.00 |
Twitter has already become an important facet of the social Web, allowing users and organizations to share their ideas and subscribe to others' updates. Notably, it has been actively used during political events such as the recent elections in Iran, Spain and the USA, or the uprisings in the Middle East, to the extent that it is claimed that revolutions will be tweeted. However such claims are yet to be empirically established. In this work we study an extensive set of millions of tweets themed around the 2011 regional and local elections in Spain and their accompanying civil unrest. Our analysis shows that mainstream political entities were overshadowed on Twitter by grassroots movements discussing alternative points of view, at both regional and national levels. By tracking temporal trends we discover an explosion of interest in the Spanish protests following the establishment of the first protest camps in Madrid. The subsequent spontaneous setting up of similar camps in almost every town in Spain was widely reported at the time, and our spatial analysis quantitatively attests that the protests were indeed widely discussed in all parts of the country. Our results suggest that online social networks serve more as media for interactive engagement between grassroots peers, rather than as platforms for politicians to attract mass attention.