Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The UN Security Council debates 1992-2023

Mirco Schoenfeld, Steffen Eckhard, Ronny Patz, Hilde van Meegdenburg, Antonio Pires

TL;DR

The paper introduces an updated UN Security Council Debates corpus covering 1992–2023 with 106,302 speeches across 6,233 meetings, providing a rich resource for text-as-data analysis of international diplomacy. It details a semi-automated pipeline that extracts speeches from public transcripts, cleans and segments the text, and labels speakers by country or affiliation, enabling fine-grained analysis of who speaks and about what. Descriptive statistics reveal rising activity, a shift toward longer and more open debates since 2014, and a notable role for UN bureaucracy alongside the P5, highlighting intricate power dynamics in UNSC discourse. The dataset offers a platform for NLP investigations into topic evolution, coalition dynamics, and the influence of international bureaucracies on security policymaking, with potential extensions to topics like Women, Peace and Security and climate-related issues.

Abstract

This paper presents an updated dataset containing 106,302 speeches held in the public meetings of the UN Security Council (UNSC) between 1992 and 2023. The dataset is based on publicly available meeting transcripts with the S/PV document symbol and includes the full substance of individual speeches as well as automatically extracted and manually corrected metadata on the speaker, the position of the speech in the sequence of speeches of a meeting, and the date of the speech. After contextualizing the dataset in recent research on the UNSC, the paper presents descriptive statistics on UNSC meetings and speeches that characterize the period covered by the dataset. Data highlight the extensive presence of the UN bureaucracy in UNSC meetings as well as an emerging trend towards more lengthy open UNSC debates. These open debates cover key issues that have emerged only during the period that is covered by the dataset, for example the debates relating to Women, Peace and Security or Climate-related Disasters. The corpus is available online: https://doi.org/10.7910/DVN/KGVSYH

The UN Security Council debates 1992-2023

TL;DR

The paper introduces an updated UN Security Council Debates corpus covering 1992–2023 with 106,302 speeches across 6,233 meetings, providing a rich resource for text-as-data analysis of international diplomacy. It details a semi-automated pipeline that extracts speeches from public transcripts, cleans and segments the text, and labels speakers by country or affiliation, enabling fine-grained analysis of who speaks and about what. Descriptive statistics reveal rising activity, a shift toward longer and more open debates since 2014, and a notable role for UN bureaucracy alongside the P5, highlighting intricate power dynamics in UNSC discourse. The dataset offers a platform for NLP investigations into topic evolution, coalition dynamics, and the influence of international bureaucracies on security policymaking, with potential extensions to topics like Women, Peace and Security and climate-related issues.

Abstract

This paper presents an updated dataset containing 106,302 speeches held in the public meetings of the UN Security Council (UNSC) between 1992 and 2023. The dataset is based on publicly available meeting transcripts with the S/PV document symbol and includes the full substance of individual speeches as well as automatically extracted and manually corrected metadata on the speaker, the position of the speech in the sequence of speeches of a meeting, and the date of the speech. After contextualizing the dataset in recent research on the UNSC, the paper presents descriptive statistics on UNSC meetings and speeches that characterize the period covered by the dataset. Data highlight the extensive presence of the UN bureaucracy in UNSC meetings as well as an emerging trend towards more lengthy open UNSC debates. These open debates cover key issues that have emerged only during the period that is covered by the dataset, for example the debates relating to Women, Peace and Security or Climate-related Disasters. The corpus is available online: https://doi.org/10.7910/DVN/KGVSYH

Paper Structure

This paper contains 9 sections, 6 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Constructing the UN Security Council Debates corpus
  3. Extracting raw text from PDF-documents:
  4. Cleaning up raw text:
  5. Splitting up raw text into distinct speeches:
  6. Labeling speeches by speakers' names and countries or affiliations:
  7. Descriptive Statistics of the UN Security Council Debates Corpus
  8. Meeting Patterns of the UNSC: From Formalized Discussions to Open Debates?
  9. Conclusions and outlook

Figures (6)

  • Figure 1: Meetings over time.
  • Figure 2: Numbers of speeches per year.
  • Figure 3: Number of meetings per agenda item and time period.
  • Figure 4: Numbers of tokens over time.
  • Figure 5: Numbers of speeches per meeting -- dots indicate the number of meetings with a certain number of speeches. Logarithmic scaling on the y-axis.
  • ...and 1 more figures