Analysis of the public debate on DDL Zan on Twitter : an application of the Structural Topic Model

Social media play a key role in analysing the public debate on political issues. The amount of information available and the ability to share opinions on various topics has led to an increase in the production of user-generated content. Indeed, data from social media platforms such as Twitter offer new possibilities because it provides a public arena to understand feedback and opinions on certain political topics. In this article, we propose an analysis of the debate on Twitter regarding the DDl Zan, the bill that aims to toughen the penalties of crimes and discrimination against homosexuals, transsexuals, women, and disabled people. We define a strategy to extract the main topics of the public debate considering the role assumed by the so-called influencers within the Twitter community. for this reason, we chose to use the Structural Topic Model (STM) including in the model the covariates describing the user's profiles of interest. This allows to identify the most relevant topics and to assess the impact of covariates in the model to study how covariates impact on them.