top of page

THE U.N. SECURITY COUNCIL DEBATES PLATFORM

An Automated Text Analysis of Diplomatic Speeches

heatmap ed.png

What is it?

  • An online open corpus containing the more than 80 thousand speeches delivered at the United Nations Security Council (UNSC) between 1995 and 2020. The next update will comprise all speeches delivered from 1992 and 2022.

  • An interactive tool that allows users to navigate through the corpus texts and its metadata, visualize the frequency of meetings and topics, and the outcomes of selected analysis of the speeches, such as keywords, sentiment, and similarity between members. It is possible to apply filters for member, topic, and time span.

Why is it significant?

  • Previous research on the UNSC relies on case studies, usually focusing on specific members or conflicts. The dataset makes it possible to study the Council as a whole, covering all of its members and all debated conflicts. Also, it contains a comprehensive set of variables providing a deeper look into the characteristics of each meeting, including the content of all speeches, complementing the information provided by the UN website and its reports.

Who will benefit from it?

  • The UNSC and UN Secretariat will have at their disposal detailed information about the Council’s practice during the whole post-Cold War period.

  • Diplomats, researchers, and UN experts interested in monitoring the practice of the UNSC and the positions expressed by its members through their speeches.

What stage is it at?

  • The process of collecting speeches for the new version of the dataset is complete. We already performed sentiment analysis in the corpus and ran descriptive analysis to observe the frequency of meetings by a preliminary version of the topics variable. It’s necessary to standardize topics’ labels and build the online platform to present all of the analyses.

FIG3.png

ACCESS THE DATA

Access the original dataset by M Schoenfeld, S Eckhard, R Patz, H Meegdenburg, A Pires.

SENTIMENT AND FREQUENCY

Explore the frequency of themes and speaker sentiment through a heatmap

ACCESS THE PROTOTYPE

Test the tool for a sample of UN Security Council members

Antonio Henrique Pires dos Santos

Co-author of the dataset

Chairman of the DiploData startup.

Ph. D. in International Relations at the Federal University of Pernambuco (UFPE)

Member of the research groups Núcleo de Estudos em Política Comparada e Relações Internacionais (NEPI-UFPE) and Computational and Experimental Political Science Lab (CPCEX Lab-UFPE).

foto antonio delbrasonu.jpg

Milestones and progress

2019

2020

2021

2022

2023

M Schönfeld (Univ. of Bayreuth), S Eckhard (Zeppelin Univ.), R Patz (Hertie School of Govt.), and H Meegdenburg (Leiden Univ.) release the first version of the UN Security Council Debates, with 65,393 UNSC speeches between 1995 and 2017 (release article published at The Journal of European Public Policy)

PhD candidate A Pires (UFPE) joins as a co-author following his contribution with data-cleaning, inclusion of new variables, and collection of new speeches.

Expansion of the corpus to 1995-2020, totaling 82,165 speeches.

Funding via the CNPq grant “Chamada Universal” as part of the research project “Multilateralism and global challenges: past, present, and future” (420041/2021-4). Presentation of the tool to UNITAR and to the UN Library.

Expansion of the dataset to 1992-2022.

bottom of page