Social Media Analysis during Political Turbulence, PLoS ONE, 2017

An in-depth exploratory analysis of Twitter posts for two major political events, a referendum and the follow-up elections revealed hidden aspects of public sentiment towards political actors. This work combined five types of analysis: volume analysis, entity identification, sentiment detection, sarcasm correction and topic analysis. Main findings include the co-enhancement of results between the types of analysis and the quantitative aspects of data analytics visualized for social and political interpretations. Worries, fears and other sentiments provide insight to the mindset of the citizens as they were called upon to decide on their future. read more