Advancing Text-as-Data Analysis of International Organizations

Abstract

Text-as-data research is quickly expanding in political science and International Relations, including in the study of International Organizations (IOs). Two major forces are reshaping this agenda. One is the volume of relevant text that is readily accessible for IO-focused analysis. The second is the unprecedented expansion of the set of high-powered text-analytical tools. This includes – but is not limited to – large embeddings-based language models, such as BERT, and various Machine Learning (ML) techniques. This panel brings together contributions that use text-as-data techniques to study how IOs are talked about (e.g., in policy-relevant documents, in traditional and social media, by political leaders), what IOs say and write (their textual output), and other related important questions. It thus serves the community of IO scholars using these novel analytical tools and data as a venue for exchange, as well as IO and political science scholars at large for exploring the potential (and pitfalls) of this rapidly advancing methodology.

Title	Details
Speech is silver, but resolutions are gold: Reassessing member state agency in United Nations decision-making (1990-2019)	View Paper Details
International recognition of European Union ‘actorness’: Language-based evidence from United Nations General Assembly speeches 1970-2020	View Paper Details
Text analysis of international organizations: A new framework to leverage large-scale data	View Paper Details
Media Reporting on International Organizations: Using Machine Learning to Identify Worldwide Patterns	View Paper Details