PhD Thesis Prototype — Dominique S. Loyer (UQAM)
Citation Key: loyerModelingHybridSystem2025
Note
Version stable : v2.4 (22 février 2026) — Restructuration consolidée
- Fact-Checking multi-sources (Google Fact Check API)
- E-E-A-T (Experience, Expertise, Authority, Trust)
- NER — Extraction d'entités nommées (spaCy)
- GraphRAG — Réseau Neuro-Symbolique (D3.js)
- IR Engine — BM25, TF-IDF, PageRank
- TREC AP88-90 — 242,918 documents (corpus complet)
- Métriques — Precision, Recall, nDCG, MRR
- Bias Analysis — Détection de biais
systemFactChecking_Sandbox/
├── syscred/ # ← Package Python unique et propre
│ ├── __init__.py
│ ├── backend_app.py # API Flask
│ ├── verification_system.py # Système principal
│ ├── config.py
│ ├── ner_analyzer.py # ← À restaurer
│ ├── eeat_calculator.py # ← À restaurer
│ ├── graph_rag.py
│ ├── ontology_manager.py
│ ├── api_clients.py
│ ├── seo_analyzer.py
│ ├── ir_engine.py
│ ├── eval_metrics.py
│ ├── trec_retriever.py
│ ├── trec_dataset.py
│ ├── liar_dataset.py
│ ├── database.py
│ └── static/
│ └── index.html
├── huggingface_space/
│ ├── Dockerfile # ← Mis à jour
│ └── README.md
├── requirements.txt # ← Allégé pour Render
├── requirements-full.txt # ← Version complète pour HF/local
├── Dockerfile # Pour Render
├── .env
├── README.md
├── 03_Docs/
├── 99_Archive/ # ← Anciennes versions archivées
└── ...A neuro-symbolic AI system for verifying information credibility that combines:
- Symbolic AI: Rule-based reasoning with OWL ontologies (RDF/Turtle)
- Neural AI: Transformer models for sentiment analysis (DistilBERT) and NER (spaCy)
- IR Engine: BM25, TF-IDF, and PageRank estimation
- E-E-A-T: Google Quality Rater Guidelines scoring
The system provides explainable credibility scores (High/Medium/Low) with detailed factor breakdown.
systemFactChecking/
├── README.md # This file
├── LICENSE
├── pyproject.toml # Python/PyPI config
├── Dockerfile # Render deployment (Lite, <528MB)
├── Dockerfile.huggingface # HuggingFace Spaces (Full, with ML)
├── requirements.txt # Standard dependencies
├── requirements-lite.txt # Render (no ML models)
├── requirements-full.txt # HF/Local (PyTorch, spaCy, etc.)
├── .env # Environment variables
├── .gitignore
│
├── syscred/ # ⭐ SOURCE CODE (single source of truth)
│ ├── __init__.py
│ ├── backend_app.py # Flask server (entry point)
│ ├── verification_system.py # Main verification pipeline
│ ├── api_clients.py # External API clients
│ ├── config.py # Configuration
│ ├── database.py # DB management (Supabase/PostgreSQL)
│ ├── db_store.py # Database storage layer
│ ├── ir_engine.py # IR engine (BM25, TF-IDF)
│ ├── ontology_manager.py # Ontology management (OWL/RDF)
│ ├── seo_analyzer.py # SEO analysis
│ ├── eval_metrics.py # Evaluation metrics (MAP, NDCG, etc.)
│ ├── graph_rag.py # GraphRAG module (D3.js)
│ ├── ner_analyzer.py # NER (spaCy)
│ ├── eeat_calculator.py # E-E-A-T scoring
│ ├── trec_retriever.py # TREC evidence retrieval
│ ├── trec_dataset.py # TREC AP88-90 corpus parser
│ ├── liar_dataset.py # LIAR benchmark dataset
│ └── static/
│ └── index.html # Frontend dashboard
│
├── ontology/ # OWL/RDF ontology files
│ ├── sysCRED_onto26avrtil.ttl # Base ontology schema
│ └── sysCRED_data.ttl # Ontology data (triplets)
│
├── huggingface_space/ # HuggingFace Space config
│ ├── Dockerfile
│ └── README.md
│
├── tests/ # Unit tests
├── 01_Presentations/ # PhD presentations (LaTeX)
├── 03_Docs/ # Documentation & dev logs
├── 04_Bibliography/ # Research references
├── 99_Archive/ # Archived older versions
└── assets/ # Images and graphs
Paramètres clés:
| Paramètre | Description | Valeur par défaut |
|---|---|---|
HOST |
Adresse du serveur | 0.0.0.0 |
PORT |
Port du serveur | 5000 |
DEBUG |
Mode debug | true |
LOAD_ML_MODELS |
Charger les modèles ML | true |
WEB_FETCH_TIMEOUT |
Timeout HTTP (sec) | 10 |
cd systemFactChecking
python -m venv venv
source venv/bin/activate
pip install -r requirements-full.txt
python -m syscred.backend_app
# → Access at http://localhost:5001docker build -t syscred-lite -f Dockerfile .
docker run -p 5000:5000 --env-file .env syscred-litedocker build -t syscred-full -f Dockerfile.huggingface .
docker run -p 7860:7860 --env-file .env syscred-full| Endpoint | Method | Description |
|---|---|---|
/api/verify |
POST | Full credibility verification |
/api/seo |
POST | SEO analysis only |
/api/ontology/stats |
GET | Ontology statistics |
/api/ontology/graph |
GET | D3.js graph data |
/api/trec/search |
POST | TREC evidence retrieval |
/api/trec/metrics |
POST | Calculate IR metrics |
/api/trec/corpus |
GET | TREC corpus info |
/api/trec/health |
GET | TREC module health |
/api/health |
GET | Server health check |
curl -X POST http://localhost:5001/api/verify \
-H "Content-Type: application/json" \
-d '{"input_data": "La Terre est ronde"}'| Factor | Weight | Description |
|---|---|---|
| Source Reputation | 25% | Known credible sources database |
| Domain Age | 10% | WHOIS lookup for domain history |
| Sentiment Neutrality | 15% | Extreme sentiment = lower score |
| Entity Presence | 15% | Named entities (ORG, PER) |
| Text Coherence | 15% | Vocabulary diversity |
| Fact Check | 20% | Google Fact Check API results |
E-E-A-T Score (Google Quality Rater):
- Experience: Domain age, content richness
- Expertise: Technical vocabulary, citations
- Authority: Estimated PageRank, backlinks
- Trust: HTTPS, unbiased sentiment
Pondérations des scores:
SCORE_WEIGHTS = {
'source_reputation': 0.25, # Réputation de la source
'domain_age': 0.10, # Âge du domaine
'sentiment_neutrality': 0.15, # Neutralité du ton
'entity_presence': 0.15, # Présence d'entités vérifiables
'coherence': 0.15, # Cohérence textuelle
'fact_check': 0.20 # Résultats fact-check
}# Google Fact Check API key
export SYSCRED_GOOGLE_API_KEY=your_key_here
# Supabase Database
export SYSCRED_DATABASE_URL=postgresql://...
# Base URL
export SYSCRED_BASE_URL=https://syscred.uqam.ca
# Server settings
export SYSCRED_PORT=5001
export SYSCRED_DEBUG=true
export SYSCRED_ENV=productionVariables d'environnement:
export SYSCRED_ENV=production # Environnement (dev/prod/testing)
export SYSCRED_PORT=8080 # Port personnalisé
export SYSCRED_GOOGLE_API_KEY=xxx # Clé Google Fact Check
export SYSCRED_LOAD_ML=false # Désactiver ML| Platform | URL | Type |
|---|---|---|
| HuggingFace | DomLoyer/syscred | Full (ML models) |
| Render | syscred-deploy-v2 | Lite (no ML) |
| UQAM | syscred.uqam.ca | Mirror (iframe → HF) |
@software{loyer2026syscred,
author = {Loyer, Dominique S.},
title = {SysCRED: Neuro-Symbolic System for Information Credibility Verification},
year = {2026},
publisher = {GitHub},
url = {https://github.com/DominiqueLoyer/systemFactChecking}
}MIT License — See LICENSE.
| Version | Date | Changes |
|---|---|---|
| v2.4 | Feb 22, 2026 | Consolidated restructuration, full TREC, bias analysis |
| v2.3 | Feb 8-9, 2026 | NER, E-E-A-T, Google Fact Check API, GraphRAG, blue glow UI |
| v2.2 | Jan 29, 2026 | GraphRAG, D3.js graph, Docker & Supabase integration |
| v2.0 | Jan 2026 | Complete rewrite, modular architecture, REST API |
| v1.0 | Apr 2025 | Initial prototype |
