Europe’s Digital Political Discourse: Mapping Narratives with a Multi-Platform, Multilingual Dataset
摘要
The ActEU Digital Political Discourse Dataset advances the study of online political communication by addressing critical gaps in existing resources. It provides multilingual (nine European languages), cross-platform (Twitter/X, Telegram, news outlets, institutional websites), and actor-centered data across ten countries. A sophisticated hybrid annotation method (combining human expertise and machine learning) enables detailed semantic labeling across topics such as climate change, migration, and gender equality, with annotations for stance, tone, populist rhetoric, and political trust. Despite known limitations in classifier performance and temporal coverage, the dataset offers academics, civil society, and regulators a robust tool for understanding evolving online political discourse.