Aller au contenu

Données ouvertes

Bibliothèque

Jeux de données ouverts, entièrement documentés — interrogeables ici, et lisibles par n’importe quel LLM.

Les titres et les descriptions proviennent des sources de données, en anglais.

2 jeux de données

public-health

  • Health Burden Signals (derived)

    Global health burden signals (obesity trends, PM2.5 exposure, anomalies)

    Annual ML-enriched population-health signals for ~190 territories, derived from WHO Global Health Observatory data: adult obesity prevalence by sex, the 5-year obesity trend, annual mean PM2.5 concentration and the PM2.5 exceedance multiple versus the WHO 15 µg/m³ guideline, with 30-period change volatility, 3-year momentum, 3-sigma anomaly flags versus a trailing 10-year baseline, naive-drift 1-year forecasts, and a per-year cross-country volatility rank. Raw data: World Health Organization (non-commercial). All rows carry ISO alpha-3 country_code and join cleanly with other health datasets.

    • obesity
    • air-pollution
    • pm25
    • public-health
    lignes
    40 397
    Qualité
    96
    Mis à jour
    22 sept. 2026
    À jour
    Licence
    Usage commercial interdit

  • openFDA drug safety intelligence (agent-curated)

    openFDA drug safety signal intelligence (quarterly)

    Quarterly adverse-event reporting signals for the 40 most-reported drugs in the FDA Adverse Event Reporting System (FAERS), from the official keyless openFDA drug/event API. Per drug: report counts for the trailing 91-day window ending at openFDA's last_updated, serious-outcome and death-outcome counts and shares, quarter-over-quarter report change vs the prior 91-day window, and a documented 0-100 reporting-signal composite (35% min-max(log1p(volume)) + 35% min-max(serious share) + 30% min-max(clipped positive trend)) ranked with deterministic tie-breaks. Only aggregate counts are stored — no individual reports, no patient data. Caveats: FAERS is spontaneous reporting, not incidence — no prescription denominator, Weber effect and media attention bias volumes; the score is a reporting-burden triage signal, not a safety ranking, and must never drive medical decisions. Primary key: drug_term. Cadence: quarterly.

    • health
    • signals
    • quarterly
    • rankings
    lignes
    40
    Qualité
    100
    Mis à jour
    25 sept. 2026
    À jour
    Licence
    Usage commercial OK

Utiliser votre propre clé d’IA

Une fois l’allocation gratuite du jour épuisée, les fonctions d’IA peuvent passer par votre propre compte fournisseur.

Conservée uniquement dans cet onglet (effacée à sa fermeture) et envoyée avec chaque requête d’IA. Nos serveurs l’utilisent pour cette requête et ne la stockent ni ne la journalisent jamais.