AUTOMATING THE A NALYSIS OF I NSTITUTIONAL D ESIGN IN I NTERNATIONAL AGREEMENTS APREPRINT Anna Wróblewska 1 , Bartosz Pieli ´ nski 2 , Karolina Seweryn 1,3 , Sylwia Sysko-Roma ´ nczuk 4 , Karol Saputa 1 , Aleksandra Wichrowska 1 , Hanna Schreiber 2 , 1 Faculty of Mathematics and Information Science, Warsaw University of Technology, Warsaw, Poland Email: anna.wroblewska1@pw.edu.pl 2 Faculty of Political Science and International Studies, University of Warsaw, Warsaw, Poland Email: b.pielinski@uw.edu.pl 3 NASK - National Research Institute, Warsaw, Poland Email: karolina.seweryn@nask.pl 4 Faculty of Management, Warsaw University of Technology, Warsaw, Poland ABSTRACT This paper explores the automatic knowledge extraction of formal institutional design - norms, rules, and actors - from international agreements. The focus was to analyze the relationship between the visibility and centrality of actors in the formal institutional design in regulating critical aspects of cultural heritage relations. The developed tool utilizes techniques such as collecting legal documents, annotating them with Institutional Grammar, and using graph analysis to explore the formal institu- tional design. The system was tested against the 2003 UNESCO Convention for the Safeguarding of the Intangible Cultural Heritage. Keywords Natural Language Processing · Information Extraction · Institutional Design · Institutional Grammar · Entity Graph · UNESCO · Intangible Cultural Heritage 1 Introduction Solving contemporary interlinked and complex problems requires international cooperation. This cooperation is expressed through international agreements establishing norms, rules, and actors to facilitate collaboration between nations. Though digitization processes have facilitated access to official documents, the sheer volume of international agreements makes it challenging to keep up with the number of changes and understand their implications. The formal institutional design expressed in legal texts is not easily comprehensible. This is where computational diplomacy with NLP models comes into play. They hold the potential to analyze vast amounts of text effectively, but the challenge lies in annotating legal language Libal (2020). Our paper and accompanying prototype tool aims to facilitate the analysis of institutional design in international agreements. The authors annotated the legal text using Institutional Grammar (IG), a widely used method for extracting information from written documents Frantz and Siddiki (2021). IG offers a comprehensive analysis of the rules that regulate a particular policy, revealing the relationships between rules, reducing complexity, and identifying key actors involved in the policy. Despite being relatively new, IG has achieved high standardization Frantz and Siddiki (2021) through close collaboration between political/international relations and computer scientists. As such, the prototype aligns well with the intersection of these two areas of scientific research. The growing popularity of IG in public policy studies has led to the development of several tools helping to upscale the usage of IG. There is the IG Parser created by Christopher Frantz 1 , the INA Editor created by Amineh Ghorbani (https://ina-editor.tpm.tudelft.nl/), and attempts have been made to develop software for the automatization of IG annotations Rice et al. (2021). The first tool is handy in manually annotating legal texts, and the second is 1 https://s.ntnu.no/ig-parser-visual arXiv:2305.16750v1 [cs.CL] 26 May 2023