Text mining e social media: Anatomia di una crisi di governo

In August 2019, when Matteo Salvini asked to return to the ballot box, the Italian government entered into crisis. This generated a large debate that also took place on social media – particularly on Twitter, which has now become increasingly important as a place for political debate. This study presents the analysis of the Twitter discourse related to government crisis, in the period following the resignation of the Prime Minister. Text mining and social network analysis techniques have been integrated to study the positioning of political leaders and the topics of major interest. In particular, we used the Emotional Text Mining (ETM) and Semantic Brand Score (SBS) techniques. The SBS served as an indicator of the most relevant topics in the discourse. It was used to determine the importance of the themes emerging from the EMT. This importance was measured along the three dimensions of prevalence, diversity, and connectivity – considering textual association patterns and cooccurrence networks. The integration of the two procedures allowed to describe the public perception of the crisis and to identify the symbolic-cultural matrix and the sentiment related to each discourse topic.