Redefining the Configuration of Public Opinion in the Digital Era: A Computational Analysis of Media and Social Dynamics

Análise Computacional de Dinâmicas Midiáticas e Sociais

Abstract

This study analyzes the relevance of classical communication theories ─agenda setting and social representations─ and highlights their continued applicability in explaining the behavior of mass audiences in the digital environment. The research proposes a methodological approach based on Bayesian inference that allows quantifying the spread of information and segmenting user groups affected by a given communication campaign. This procedure is applied to the computational analysis of 40 news articles and 61,840 tweets related to a controversial media case, demonstrating how press content can catalyze large-scale digital discussions, fuel the spread of biased narratives, and shape the formation of opinion-based communities. The findings reveal the dual role of traditional media and digital platforms: on the one hand, they focus public attention on certain issues; on the other hand, they amplify the reach and intensity of these narratives through viral dissemination. This dynamic fosters the emergence of polarized communities, often driven by overexposure to unverified content. The study underscores the urgent need to implement effective strategies and countermeasures to mitigate the effects of digital media polarization. In this regard, it proposes analytical tools to detect patterns of manipulation, identify vulnerable communities, and promote a more informed, critical, and participatory public