The goal of this paper is to evaluate the differences of behaviour of users in Twitter before and after the Spanish 2020 confinement, that went from 13th March 2020 to May 2020. The dataset we'll analyse will be tweets from 18.000 Spanish twitter profiles, categorized in 3 types: politicians (about 300 profiles), journalists (800 profiles) and random profiles (over 16.000).
First we'll try to view from a global perspective the data as well as explaining how we gathered it. Then we'll create a Bayesian model as simple as we can for each user, in order to measure the impact of confinement in user's activity. Finally we'll try to unveil the differences of the fitted model among the groups of users and study their characteristics.