🛒 FairCart AI – Bias-Aware Product Ranking

---

🚀Live Demo

🔗 Try FairCart AI here:

https://faircart-ai-2djleyssiqbm6q9kgvib5f.streamlit.app/

FairCart AI is a fairness-aware product ranking system that detects and reduces demographic bias in e-commerce recommendations while preserving ranking relevance.

This project demonstrates how algorithmic fairness techniques can be applied to real-world ranking systems such as online marketplaces.

---

🚀 Features

• 📊 Compares baseline ranking vs fairness-aware ranking
• ⚖️ Detects demographic bias (Men vs Women representation)
• 📉 Computes fairness gap and bias reduction
• 🧠 Preserves relevance when no bias is detected
• 🖥️ Interactive Streamlit UI
• 🔄 Accepts dynamic JSON product input

---

🏗️ Tech Stack

• Frontend: Streamlit
• Backend Logic: Python
• Data Handling: Pandas
• Fairness Metrics: Statistical Parity Difference
• Deployment: Streamlit Cloud

---

📂 Project Structure

faircart-ai/
│
├── src/
│ ├── fair_ranking.py # Fairness-aware ranking logic
│ ├── metrics.py # Fairness metrics
│
├── data/ # Sample datasets
├── notebooks/ # Data exploration notebooks
├── outputs/ # Saved outputs (JSON)
│
├── streamlit_app.py # Main Streamlit app
├── requirements.txt # Dependencies
├── README.md # Project documentation

---

▶️ How to Run Locally

pip install -r requirements.txt
streamlit run streamlit_app.py


🧪 Sample Input (JSON)
{
  "products": [
    {"title": "Women Casual Shoes", "maincateg": "Women", "baseline_score": 0.92},
    {"title": "Men Formal Shoes", "maincateg": "Men", "baseline_score": 0.95},
    {"title": "Women Sports Shoes", "maincateg": "Women", "baseline_score": 0.85}
  ]
}

---

📊 Fairness Metrics Explained

Fairness Gap (%) Absolute difference between men and women representation

Statistical Parity Difference (SPD) Measures bias reduction after applying fair ranking

Lower values indicate better fairness.

---

👩‍💻 Author

Pragati Gupta

-Mathematics & Computing Student

-Project built for learning and showcasing fairness-aware AI systems.

---

⭐ Future Improvements

-Support for multiple protected attributes

-Real-world datasets integration

-Advanced fairness constraints

-Model-based ranking

---

📜 License

This project is open-source and free to use for learning and research.