CERNO Political
See how issues move before they become headlines.
An issue radar for Indonesia: emerging public issues detected against each region's own baseline, attributed to real administrative geography, and backed by evidence — with confidence stated, not implied.
- 38
- 514
- 7,285
What this surface does
Emerging issue detection
Volume, velocity, and persistence measured against per-region baselines — abnormality is computed, not felt.
Geographic attribution
Signals resolve to province, regency, and district on a decree-verified gazetteer, with disambiguation for place names that double as common words.
Issue lifecycle
Discovered → emerging → rising → viral → critical → stabilizing → resolved — with the full history kept, hour by hour.
Evidence & claims
Claims carry factual status (verified, disputed, rumor…) and link to evidence; counter-evidence is preserved.
Integrity signals
Suspicious amplification is partitioned from organic volume — never silently blended, never used to rewrite geography.
Forecast calibration
The system scores its own forecasts against outcomes — and when a model loses to the base rate, that result is kept and constrains its use.
Where this actually stands
EVALUATION — SYNTHETIC-FED
The pipeline runs end-to-end on real infrastructure (PostgreSQL + object storage) with production-grade tests, but is currently fed by synthetic staging worlds. No live source is connected yet; nothing on this page is live public data.
One part of one system.
CERNO surfaces share memory, evidence, confidence, decisions and outcomes. What one learns is available to the others — that shared state is the asset, not any single view.