GDELT Plus
International Interaction Visualizations
Get Started!

Introduction

Gigabytes of data flowing in daily, drawn from all over the world, across print, broadcast and web sources, in over 100 different languages

Entities

Google-curated algorithms and crawlers pulling out names of countries, organizations, religions, ethnicities, ...

Actions

Algorithms extracting keyword patterns to classify interactions by verbal/material and cooperation/conflict, as well as a three-level hierarchy of characteristics and impact scores.

Attitude

Measures of press coverage in terms of sources, articles and mentions, as well as AI model-driven extraction of tone of coverage.

Event Flow

Sankey Chart visualizing the density of bidirectional interaction among countries in the world

Perception Radar

Starplot to indicate how major countries view the world's superpowers

Interaction Networks

Cluster visualization of mutual interaction reveals cliques and foci of interactivity among countries

V1: Action Flows

A visualization of the relative density of interaction among countries, from a set of agent countries to a set of target countries

Agent Countries
Target Countries

V2: Starplot

How Several Major Countries View the World's Superpowers.

United States

Key Patterns

  • Small radar graph indicates that America is viewed as less friendly by other countries
  • Large radar graph indicates that China has good relations with European countries.
  • Similar graphs indicate that Russia and China are quite aligned.
  • Japan and US don't seem to view other countries quite nicely.
*The metric is the Goldstein score (seen on axis of Radar chart). We compute the average Goldstein score of news sent from seven countries to China, United States, and Russia. Higher scores indicate that countries are aligned. For country's relationship to itself, we use the mean received from all countries.

V3: Interaction Network

Visualization of density of interaction among countries as a weighted network

Countries

Key Patterns

  • Interaction among OECD+ countries centers around the US
  • Extremely high interaction between US and UK, and US and China
  • Very low degree of interaction between Russia and China.
  • Israel shows highest international interaction outside Security Council countries.
*The metric is the product of the bilateral interaction among countries scaled relatively among the interaction products of all the countries in the chosen network.




VX: Reciprocity Map

View bilateral reciprocity of interactions, exploring the balance of agency and receptivity of interaction





Other Visualizations

Nimbly navigate the visualizations below.

Click images to explore!

Credits

CS 171 Visualization / 06472: Visualization (at SEAS) (SCI 0647200)

Harvard John A. Paulson School of Engineering and Applied Sciences

Dimiter Radonov

Harvard College

Sohan Dsouza

Media Lab, MIT

Zihao Wang

Harvard College

NJ Namju Lee

Harvard GSD

Reference

Data : The GDELT Project

Language: HTML, CSS, Javascript, python

Library: D3, JQuery, MapboxGL, Bootstrap