A shared hub for coordinated response to health misinformation
EU Fact Force is a collaborative platform developed by EUPHA (European Public Health Association) with support from Data For Good volunteers. The platform empowers public health professionals to counter health misinformation by:
- Connecting scientific evidence with disinformation narratives
- Visualizing knowledge graphs of research articles, claims, and concepts
- Tracking disinformation trends through integration with PGP (The Public Good Projects) monitoring data
- Enabling rapid response with validated counter-narratives
Marie, a health communicator at a national public health association, sees a viral post claiming "vaccines cause autism." She needs to respond quickly with solid evidence.
She searches "vaccines autism" on EU Fact Force and immediately sees:
- An interactive graph showing 15+ peer-reviewed articles that refute this claim
- The scientific consensus: "Refuted with high confidence"
- Current disinformation trends: 1,200 mentions this week, peak in France/Belgium
- Key evidence to cite in her response
Time to find relevant evidence: <30 seconds
- Semantic Search (FR/EN): Find relevant scientific articles even without exact keyword matches
- Interactive Knowledge Graph: Explore connections between articles, claims, and narratives
- Disinformation Trends: Visualize PGP monitoring data (volume, geography, examples)
- Researcher Upload: Members can upload scientific articles with auto-extracted metadata
- Multilingual: Interface and search in French and English
- Access Control: Authentication system for EUPHA members
- 3 priority narratives (e.g., vaccines-autism, moderate alcohol benefits, COVID misinformation)
- 100-150 scientific articles from EJPH, WHO, ECDC, and other trusted sources
- 20-30 claims per narrative (confirmed/refuted/nuanced)
- Integration with PGP data (weekly batch updates)
14_EUFactForce/
├── .github/
│ └── workflows/
│ ├── d4g-utils.yml
│ └── pre-commit.yaml
├── docs/
├── eu_fact_force/
│ ├── exploration/ # code to keep track of benchmarks
│ ├── ingestion/ # ingestion and indexing of documents
│ └── app/ # web app
├── tests/
Ce projet utilise uv pour la gestion des dépendances Python. Il est préréquis pour l'installation de ce projet.
Une fois installé, il suffit de lancer la commande suivante pour installer la version de Python adéquate, créer un environnement virtuel et installer les dépendances du projet.
uv sync
uv run pre-commit installPour exécuter le pipeline d'ingestion avec parsing Docling, installez aussi les dépendances de parsing :
uv sync --group parsingPour exécuter l'étape d'embedding dans le pipeline d'ingestion :
- le modèle utilisé est
intfloat/multilingual-e5-base sentence-transformersest requis (installé viauv sync)- prévoir plus de RAM/temps au premier chargement du modèle
A l'usage, si vous utilisez VSCode, l'environnement virtuel sera automatiquement activé lorsque vous ouvrirez le projet. Sinon, il suffit de l'activer manuellement avec la commande suivante :
source .venv/bin/activateuv run pre-commit run --all-filesUtiliser pytest pour tester votre code
uv run pytestL'application se compose d'un serveur Django, d'une base PostgreSQL (avec pgvector) et de LocalStack pour le stockage S3. Pour déployer et utiliser l'application en local :
1. Prérequis
- Python 3.12+ et uv
- Docker et Docker Compose (pour Postgres et LocalStack)
2. Variables d'environnement
Copiez le fichier d'exemple et adaptez les valeurs si besoin :
cp .env.template .envPour un usage local avec les services Docker, les valeurs par défaut de .env.template (notamment DATABASE_URL=postgresql://eu_fact_force:eu_fact_force@localhost:5432/eu_fact_force) conviennent.
3. Lancer les services (Postgres et LocalStack)
À la racine du projet :
docker compose up -dCela démarre PostgreSQL (port 5432) et LocalStack S3 (port 4566). Le bucket configuré est créé automatiquement au démarrage de LocalStack.
4. Installer les dépendances et appliquer les migrations
uv sync
uv run python manage.py migrate5. (Optionnel) Créer un superutilisateur
Pour accéder à l'interface d'administration Django :
uv run python manage.py createsuperuser6. Démarrer le serveur Django
uv run python manage.py runserverL'application est alors disponible sur http://127.0.0.1:8000/. L'admin Django est accessible à http://127.0.0.1:8000/admin/ si un superutilisateur a été créé.
Utilisation du stockage S3 en local
Pour que Django utilise LocalStack pour le stockage des fichiers, décommentez et renseignez dans .env les variables S3 (voir .env.template), par exemple :
USE_LOCAL_STACK=1
AWS_ACCESS_KEY_ID=test
AWS_SECRET_ACCESS_KEY=test
AWS_STORAGE_BUCKET_NAME=eu-fact-force-files
AWS_S3_REGION_NAME=eu-west-1Sans ces variables, l'application utilise le stockage fichier local par défaut.