🤖 GraphRAG POPCORN Q&A Interface

Population Health Modelling Consensus Reporting Network (POPCORN)

A powerful question-answering interface powered by Neo4j, OpenAI, and Streamlit, built using the neo4j-graphrag framework. This app enables natural language queries over a Neo4j knowledge graph, providing contextual answers along with interactive graph visualizations.

Live System: POPCORN_GraphRAG App

---

🚀 Features

• 🔍 Natural Language Querying over Neo4j graphs using OpenAI's GPT-4o
• 🧠 GraphRAG retrieval combining full-text and vector search (HybridCypherRetriever)
• 📎 Contextual Answers generated via customizable RAG templates
• 🌐 Interactive Graph Visualization with PyVis and NetworkX
• 🎛 Advanced Filter Controls for refining visualizations by node and edge type
• 🔐 Google OAuth Authentication for secure user access
• 📚 Expandable Context and detailed JSON Debugging information for transparency

---

📦 Installation

1. Clone the repository

git clone https://github.com/your-username/graphrag-streamlit-app.git
cd graphrag-streamlit-app

2. Set up a Python environment

python3 -m venv venv
source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Configure environment variables

Create a .env file in the root directory containing:

NEO4J_URI=bolt://<your-neo4j-host>:7687
NEO4J_USERNAME=neo4j
NEO4J_PASSWORD=your_password
OPENAI_API_KEY=sk-...
GOOGLE_CLIENT_ID=your_google_client_id
GOOGLE_CLIENT_SECRET=your_google_client_secret
REDIRECT_URI=http://localhost:8501

Note: Never commit your .env file to version control.

---

▶️ Run the App

streamlit run app.py

• Access locally: http://localhost:8501
• Access remotely: http://<your-ec2-ip>:8501 (ensure port 8501 is open in your security group)

You can also disable CORS if needed:

streamlit run app.py --server.port 8501 --server.enableCORS false

---

📊 Example Use Cases

• Summarize and explore research topics like COVID-19, Brain-Heart Axis, or Cancer
• Dynamically explore paper-author-abstract relationships
• Visualize specific graph neighborhoods and relationships
• Facilitate knowledge-driven Q&A over complex health and scientific domains

---

📁 Project Structure

📦 graphrag-streamlit-app
├── app.py                 # Main Streamlit app
├── utils                  # Helper modules
│   ├── auth.py            # OAuth authentication logic
│   ├── neo4j_setup.py     # Neo4j and LLM configuration
│   └── graph_viz.py       # Graph visualization utilities
├── .env                   # API and DB credentials (ignored by Git)
├── requirements.txt       # Python dependencies
└── README.md              # Project documentation

---

🛡 Security Notes

• Protect your .env file from public exposure
• Restrict Neo4j database access appropriately
• Deploy with HTTPS and authentication mechanisms for production environments

---

🧠 Built With

• Neo4j GraphRAG
• Streamlit
• OpenAI GPT-4o
• NetworkX
• PyVis
• Streamlit OAuth

---

👤 Author

Pouria Mortezaagha
Data Analyst • AI Researcher • Full-Stack Developer
LinkedIn
✉️ [email protected]

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📝 How to Cite

If you use the GraphRAG POPCORN Q&A Interface in your research or projects, please cite:

Software

Mortezaagha, P. (2025). GraphRAG POPCORN Q&A Interface (v1.0.0). GitHub. https://github.com/pouriamrt/POPCORN

BibTeX

@software{mortezaagha_graphrag_popcorn_2025,
  author  = {Mortezaagha, Pouria},
  title   = {GraphRAG POPCORN Q\&A Interface},
  version = {1.0.0},
  year    = {2025},
  url     = {https://github.com/pouriamrt/POPCORN},
  license = {MIT},
  note    = {Streamlit + Neo4j GraphRAG app for population health modelling and knowledge graph Q\&A}
}