📧 Email Spam Classifier

A production-grade machine learning system designed to classify emails as Spam or Ham (Legitimate) with high accuracy. This project features a beautiful dark-themed web interface built with Streamlit and a robust REST API powered by FastAPI.

✨ Features

• 🚀 High Accuracy Models:

  • Ensemble Model: 97.12% accuracy (Voting Classifier with Random Forest + XGBoost + GBM)
  • Pipeline Model: 96.85% accuracy (Random Forest with optimized feature selection)
  • Random Forest: 96.50% accuracy (stand-alone implementation)

• 💻 Interactive Web Interface:

  • sleek, modern Dark Mode UI.
  • Real-time classification with probability scores.
  • Visual confidence metrics (Spam vs. Ham probability).

• 🔄 Batch Processing:

  • Analyze multiple emails simultaneously.
  • Support for bulk text input or file uploads (.txt, .csv).
  • Download detailed reports as CSV.

• 📊 Advanced Analytics:

  • Track analysis history in real-time.
  • View distribution statistics (Spam vs. Ham).
  • Filter and export historical data.

• 🔌 REST API: Full-featured API for integrating spam detection into other applications.

---

🛠️ Installation

Prerequisites

• Python 3.8 or higher
• pip (Python package manager)

Steps

1. Clone the repository

   git clone https://github.com/yourusername/email-spam-classifier.git
   cd email-spam-classifier

2. Create a virtual environment (Optional but recommended)

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # macOS/Linux
   source venv/bin/activate

3. Install dependencies

   pip install -r requirements.txt

---

🚀 Usage

1. Web Application (Streamlit)

Launch the interactive web interface:

streamlit run app.py

The app will open in your browser at http://localhost:8501.

2. REST API (FastAPI)

Start the API server:

python api.py
# OR using uvicorn directly
uvicorn api:app --reload

• API Documentation: Visit http://localhost:8000/docs for the interactive Swagger UI.
• Health Check: GET /health

---

📂 Project Structure

├── app.py                  # Main Streamlit web application
├── api.py                  # FastAPI REST API endpoints
├── src/                    # Source code for Core Logic
│   ├── predictor.py        # Model prediction logic
│   └── preprocessing.py    # Text cleaning and feature extraction
├── models/                 # Serialized ML models (joblib/pickle)
├── requirements.txt        # Project dependencies
└── README.md               # Project documentation

📊 Model Performance

Model Type	Accuracy	Features	Description
Ensemble	97.12%	576	Best for production. Combines RF, XGBoost, and GBM.
Pipeline	96.85%	576	Balanced performance using feature selection.
Random Forest	96.50%	576	Fast and reliable baseline model.

� Technologies Used

• Frontend: Streamlit
• Backend: FastAPI, Uvicorn
• Machine Learning: Scikit-Learn, XGBoost, Numpy, Pandas
• Visualization: Plotly, Matplotlib

---

📝 License

This project is open-source and available under the MIT License.