On 19 February 2026, four researchers from poltextLAB – Anna Takács, Barbara Babolcsay, Csaba Molnár and Miklós Sebők – participated in the “Methods Workshop on Large Language Models and Generative AI: Applications in Political Science, Public Policy, and Law”, supported by the V-SHIFT Lendület project and held in Bratislava.
In his presentation, The impact of committee chairmanship on successful bill introduction – a multilingual research, Csaba Molnár analysed whether opposition-led parliamentary committees influenced legislative success in Czechia, Hungary, and Slovakia between 1998 and 2023, applying multilingual text analysis supported by the Babel Machine AI tool.
In their talk, Using Un- and Semi-Supervised NLP Methods in Codebook Creation for Classification Tasks: Evidence from the ONTOLISST Project, Barbara Babolcsay and Anna Takács demonstrated how unsupervised and semi-supervised NLP methods contributed to the creation and testing of the Light Social Science Thesaurus (LiSST), improving the accuracy and efficiency of automated text classification in social science research.
