On-line Dynamics of the Far-Right and Far-Left. The Case of the September 2015 Parliamentary Election in Greece

In the midst of a polarised campaign, on September 17th the leader of Golden Dawn Nikos Michaloliakos publicly proclaimed that Golden Dawn takes political responsibility for the murder of an antifascist rapper. As much as such a provocative statement induces rage among democratic citizens, it is certain that it was a deliberate action. Whether related or not, the party augmented its vote share from 6.3% to 7% in the September 20th election. How was this provocative statement inserted into the public agenda? What sentiments did it trigger? Did it gain any salience? This paper proposes to examine short-term dynamics in public opinion, in its social media expression. The main hypothesis is that although “bigger issues covered the noise” in terms of interest for the majority of the electorate, this statement fostered reactions among and between the far right and the far left. We will search for the reactions and interactions of candidates and cadres of those two political spaces and their interactions with Twitter users. We are particularly interested in psychosocial aspects of on-line interactive extremism. By bringing together scholars from political sociology, psychology and computer science, this paper has a threefold scope. It firstly aspires to contribute to the study of radicalism. Secondly, to fill in the literature gap regarding the use of new media to track opinion dynamics, by combining data scientists’ technical skills with political scientists’ concerns. Thirdly, to touch upon methodological issues related to the location of special publics in social media research. We perform a set of social network analysis and text mining techniques (sentiment analysis, data classification, graph mining, time series analysis) in order to analyze user content and study the social interactions. Our results are based on a set of tweets posted from multiple locations in Greece during the period under study.