Powered by the GDELT Project
This map uses the GDELT Events 2.0 dataset, which automatically scans news articles from around the world every 15 minutes and extracts political events — who did what to whom, and where. Each dot on the map represents one news article in which a political event was identified.
GDELT's coverage is not geographically neutral. The dataset is dominated by English-language and Western media — primarily outlets from North America, the UK, and Australia. Regions with less English-language media presence (Central Africa, Central Asia, parts of Latin America) are significantly underrepresented. High event counts for a country may reflect media attention as much as actual activity.
Events are classified using the CAMEO coding scheme, a political science framework that only captures actor-to-actor political interactions. Non-political topics (natural disasters, economics, culture) are not covered.
The Political Landscape view colors countries by their Goldstein score — a fixed value assigned to each event type ranging from −10 (most destabilizing) to +10 (most cooperative). The map shows each country's weighted average Goldstein score as an initiating actor (Actor 1), reflecting how aggressively or cooperatively that country behaves in its outgoing interactions.