Computational and
Digital Projects

With the lead of Zoltán Kmetty, we have been using digital trace data to address both methodological and sociological questions. In our main Data Donation project, we collected the full Facebook and Google data of 750 participants alongside a comprehensive survey questionnaire.

Using computational social science methods, machine learning, and Natural Language Processing techniques, we analyzed determinants of willingness to donate digital trace data, validated experimental measures of data-sharing behavior against real-world actions, classified social position based on social media behavioral patterns, examined invisible forms of political engagement online, predicted COVID-19 vaccine uptake using digital behavioral data, and investigated the dynamics of political filter bubbles across platforms and over time.

In another recent project, we examined political content consumption and polarization on TikTok during the 2026 Hungarian parliamentary election campaign, with a particular focus on how algorithmic recommendation systems shape personalized political information environments. Using a large-scale data donation approach, participants voluntarily shared their TikTok viewing histories alongside survey data, allowing us to analyze political exposure, filter bubbles, and campaign-related information flows across the platform.