News

A new study by Frank Baumgartner, Shaun Bevan and Miklós Sebők, titled “Agenda-setting studies in public policy: Origins, development, and new possibilities for coding in the age of AI”, has been published in the journal Communication and Change. The article reviews over fifty years of research building on McCombs and Shaw’s seminal 1972 agenda-setting study, charting how this foundational work influenced both communications and political science and giving particular attention to methodological innovation. The authors highlight how increasingly powerful computing, “text-as-data” approaches, automated classification systems and emerging artificial intelligence technologies have transformed agenda-setting research, enabling larger empirical projects and new analytical possibilities.Framed by developments such as machine-learning and AI, the paper suggests future directions for the field that build on classic theory while adapting to new data environments and technological tools. The full study is available here:https://link.springer.com/epdf/10.1007/s44382-026-00021-8

The following interns join poltextLAB in the 2025-26 spring semester: During the spring term, the interns will work on political text analysis, help with collecting and cleaning data, and support the development of computational tools for social science research.

On 10 February 2026, Miklós Sebők delivered a seminar at the Department of Political and Social Sciences, University of Catania, titled The AI-Supported Analysis of Political Regimes. Co-organised by the Jean Monnet Chair EuDARe and the Piaceri CRIDEM project, the event presented findings from two AI-assisted text-analysis studies on Hungarian democracy. Dr, Sebők examined changing media attitudes towards COVID-19 vaccines and the spread of illiberal policy frames in parliamentary debates on immigration and the pandemic. He also introduced poltextLAB and its methodological contributions to computational social science.

On 28 January 2026, Miklós Sebők commented to the Hungarian news outlet 24.hu on the political implications of a fatal Immigration and Customs Enforcement (ICE) operation in Minneapolis. In his remarks, he addressed how the incident and the circulation of video footage influenced the Trump administration’s public communication. Sebők briefly outlined the institutional role of ICE and Customs and Border Protection (CBP), and highlighted potential consequences for political polarisation and public trust in the United States. Further details of the event can be found here: https://24.hu/kulfold/2026/01/28/minneapolis-trump-ice-noem-pretti/

On January 29, 2026, poltextLAB will hold a full-day training on Generative Artificial Intelligence (GenAI) in the center of Budapest. The program is designed for researchers and professionals who work with documents, writing, or data and want to understand how generative AI can be applied in their everyday work. The training will provide an introduction to the fundamentals of generative models, hands-on practice in prompt engineering and model selection, and practical examples of research applications. It will also address key ethical and legal issues, including data protection, intellectual property, and proper citation practices. More details and registration: https://bit.ly/4n3Du7E

On 12 December 2025, Miklós Sebők received the Bibó István Memorial Prize, awarded by the Hungarian Political Science Association. The prize, founded in 2014, is regarded as the most prestigious lifetime achievement award in Hungarian political science and is conferred annually on a single Hungarian scholar. In his talk, Dr Sebők presented the opportunities and challenges of using AI in political research.

As part of the Data-SHS Week 2025 Paris Île-de-France, researchers and engineers gathered to discuss the growing role of data and methods in the humanities and social sciences. The week featured conferences, workshops, and training sessions organized locally by the Paris-region University Data Platforms, in collaboration with numerous institutional partners. Within the session “Recent advances in the use of AI methods in the humanities and social sciences,” Barbara Babolcsay from poltextLAB presented work on improving the exploration and accessibility of research data through the automated assignment of concepts in research data repositories. The presentation highlighted how AI-based methods can support researchers in navigating large and complex corpora, enhancing data reuse and fostering new research perspectives in the humanities and social sciences.

On 2 December 2025, Barbara Babolcsay delivered a presentation at the 17th European DDI Users Conference as part of the panel “Challenges of Interoperability”. In her presentation, titled Using NLP Methods in Social Sciences – Experience and Opportunities, she outlined recent applications of clustering, topic modelling and automated text classification developed within the Ontolisst project, highlighting their relevance for large-scale social science research. Further details of the event can be found here: https://events.geant.org/event/1879/timetable/