On 26 June 2023, Rebeka Kiss attended the Tag der Parlamentsforschung – Day of Parliamentary Research conference organised by the Österreichisches Parlament in Vienna. The presentation entitled “Copy and paste explanatory notes? Text reuse approach with data from Hungary” gave insight into some preliminary empirical results related to her thesis.
The University of Antwerp hosted the 15th Annual CAP Conference between June 27 and 29, 2023. At the conference, Miklós Sebők presented two papers in the “Methodological Advancements” section. The two papers are entitled “Introducing the CAP BABEL MACHINE: A state-of-the-art AI solution for automated CAP coding”, and “Assessing the Policy Agenda of U.S. Presidential Speeches – A Sentence-Level Deep Learning Approach” (co-authored by Ákos Máté and Amnon Cavari). He also presented the CAP Babel Machine service (babel.poltextlab.com) of poltextLAB in the plenary session.
A new paper titled “HunEmBERT: a fine-tuned BERT-model for classifying sentiment and emotion in political communication,” authored by István Üveges and Orsolya Ring, was published in IEEE Access, a Q1 (Scimago), IF 3.467 journal in the field of computer science, on June 13, 2023. The article is available here: https://doi.org/10.1109/ACCESS.2023.3285536 The PDF version can be found here: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10149341
The journal article by Miklós Sebők, Rebeka Kiss and Ádám Kovács has been published in Parliamentary Affairs on June 12, 2023, entitled The Concept and Measurement of Legislative Backsliding. The article is available here: https://doi.org/10.1093/pa/gsad014 The PDF version can be found here: https://academic.oup.com/pa/advance-article-pdf/doi/10.1093/pa/gsad014/50576067/gsad014.pdf
This year’s call for proposals for the MTA’s Momentum Programme received 108 valid proposals, 27 in the humanities and social sciences, 40 in the life sciences and 41 in the mathematical and natural sciences. Based on the peer review, the recommendations of the Momentum Jury and the Momentum Committee, the President of the Academy decided on 31 May 2023 on the ranking of excellence and the amounts of funding to be awarded to the research teams to be supported. As a result, a new Lendület research group will be established at the Centre for Social Sciences (CSS), Budapest, led by Miklós Sebők. The five-year project, funded with over EUR 560 000, is entitled: “The changing global relations of the Visegrad countries in times of war – An artificial intelligence-assisted comparative analysis”. Dr Sebők is a research professor at the Institute of Political Science CSS, and the National Laboratory for Artificial Intelligence at CSS, as well as the principal investigator of poltextLAB (poltextlab.com), a research lab dedicated to text mining and artificial intelligence applications in the social sciences. The project will extend the lab’s machine learning-based analysis of public policy agendas to the history and current data sources of the Visegrad countries after 1990, with the help of the lab’s collaborators and national and international partners.
The European Consurtium for Political Research (ECPR) has recognized the following poltextLAB publication with an honourable mention: Miklós Sebők, Bálint György Kubik, Csaba Molnár, István Péter Járay, Anna Székely:Measuring legislative stability: a new approach with data from Hungary (You can read the journal article here.) The Jacqui Briggs EPS Prize is awarded for an article that makes a substantial contribution to the field of political science, especially articles that contribute to the understanding of new and innovative trends in political science or to innovative approaches to teaching and learning in the profession.
On 23rd May, 2023 the “Text Mining and Artificial Intelligence Training Program” was successfully held by Orsolya Ring, Ákos Máté and Péter Gelányi with 11 beginner level and 5 advanced level participants. The course introduced participants to basic text mining tasks and their applications in social sciences. They were given an overview of the differences between information retrieval and information extraction, as well as the concept of bag of words and sentiment analysis.
Members of poltextLAB attended a conference of the COMPTEXT international text mining network in Glasgow from 12 to 13 May. At the COMPTEXT conference, poltextLAB’s research was presented at the following presentations: Miklós Sebők – Amnon Cavari – Ákos Máté:Assessing the Policy Agenda of U.S. Presidential Speeches – A Sentence-Level Deep Learning Approach Ákos Máté – Miklós Sebők – Ádám Feldmann:The last frontier for automated classification in comparative politics? Using multilanguage large language models Péter Gelányi:Tipping the scales: A text mining analysis concerning the effects of government influence on online media in Hungary The conference program is available here.
The journal article by Miklós Sebők, Rebeka Kiss and István Járay has been published in the Journal of Knowledge Economy on May 8, 2023, entitled Introducing HUNCOURT: A New Open Legal Database Covering the Decisions of the Hungarian Constitutional Court for Between 1990 and 2021. The article is available here: https://doi.org/10.1007/s13132-023-01395-6 The PDF version can be found here: https://link.springer.com/content/pdf/10.1007/s13132-023-01395-6.pdf
The Visegrad Fund approved the funding of the grant proposal ‘Identifying News Slant in Crisis Communication Using Artificial Intelligence’ of the Centre for Social Sciences. Orsolya Ring, a research fellow at the Institute of Political Science and senior researcher at poltextLAB, lead the one-and-a-half-year research programme. The research, with partners from the Czech Republic, Poland and Slovakia, will investigate the bias associated with different crises in online media using artificial intelligence.