Coverage needs context
Volume alone does not prove neglect. AtlasWire needs a stable baseline across sources, languages, and places before making a coverage comparison.
Coverage calibration
Building the baseline.
AtlasWire is still measuring the difference between what happens and what receives attention. Underreporting signals will appear only after there is enough independent local and international coverage to calibrate them honestly.
No premature signals
Until that baseline is strong enough, a story is not labelled underreported. Incomplete evidence remains clearly unavailable.
Impact stays visible
While calibration continues, each event still leads to original reporting and shows the stored evidence behind its impact assessment.
Additive discovery
AtlasWire combines reviewed publisher feeds with GDELT-discovered publisher headlines. GDELT can surface reporting from any public HTTPS publisher; each publisher remains visible and enters the same event, geography, clustering, reliability, impact, and ranking pipeline.
Publisher provenance
A publisher's location describes the source, not the event. Event geography comes from headline evidence, while exact links and syndicated headlines are grouped to avoid false corroboration.
Cited AI briefs
Short event briefs are generated only from stored publisher headlines. Every sentence links to its supporting sources, numeric claims are checked against those headlines, and article bodies are not retrieved.
GDELT-discovered reporting uses data provided by the GDELT Project ↗. Headline geography uses place data provided by GeoNames ↗ under CC BY 4.0.