Dataset

global-news-event-database

GDELT, a long-running global news monitoring effort tracking events, tone and geographic references with fifteen-minute refresh cycles.

Visit provider

Dataset card

The GDELT Global Knowledge Graph is an open, real-time news event database that monitors global media in over 100 languages and extracts events, actors, and tone. It is widely used to quantify geopolitical developments at scale.

What this dataset contains

GDELT continuously ingests worldwide news articles and television transcripts, coding each into structured events that link actors, actions, and locations. The dataset captures event counts, themes, and average document tone, allowing systematic measurement of how coverage of countries, companies, and topics evolves over time. A raw licensing and file-size figure is not published alongside this catalog entry, and the underlying readme does not enumerate the full feature set.

Structure and fields

The catalog record supplied to this directory does not list specific column names, so the precise schema of the local extract is unknown. Public GDELT release files are generally distributed as delimited records with fields such as event identifier, event code, actor codes, Goldstein scale, and average tone, but these cannot be confirmed from the facts provided. Confirm the schema against the official GDELT documentation before ingesting the file into a research pipeline.

How analysts use it

Because the source content is structured news, the natural fit is event-driven and sentiment-oriented research rather than fundamental company data. Macro and emerging-market funds commonly aggregate GDELT tone and event counts to build geopolitical risk indices, monitor country-level news shocks, or stress-test portfolios against sudden policy events. Without confirmed columns or a stated coverage window in this record, any specific back-test claim would go beyond the facts available.

Documentation & references 2