This study investigates the dynamics between traditional media companies and emerging digital actors during the 2018 Brazilian Presidential Elections through a comprehensive analysis of Twitter communications. Using Latent Dirichlet Allocation (LDA) topic modeling, we analyzed 4,366,960 tweets from 892,105 users during the period of October 26–30, 2018, focusing on shared links and content from the top 20 media outlets. The research reveals that while established media companies maintained their dominant position in terms of content distribution (with 10 out of 20 top-shared domains), partisan and alternative media outlets successfully challenged their narrative. This study contributes to agenda-setting theory by quantifying how digital platforms enable alternative media to compete with traditional outlets for narrative control in political discourse, despite resource disparities.

This work is supported by the Fundação para a Ciência e Tecnologia (FCT), Portugal project grant UIDB/04466/2020.

Agenda-setting,Social media,Twitter,Topic modeling

• Computer and Information Sciences - Natural Sciences