A FastAPI-based project to detect spam emails using machine learning. This repository includes everything from training to deploying a REST API powered by a trained model.

• Python 3.8+ | dev --version: Python 3.13
• pip

pip install -r requirements.txt

uvicorn app:app --reload

Or using the Procfile (e.g., for Heroku):

web: uvicorn app:app --host=0.0.0.0 --port=${PORT:-8000}

• Dataset file: email.csv

• Contains real-world email examples labeled as ham or spam

• Mapping used in preprocessing:

  df["Category"] = df["Category"].map({"ham": 0, "spam": 1})

• Total records: [ADD_TOTAL_RECORDS]

  • 🟢 Ham: [ADD_HAM_COUNT]
  • 🔴 Spam: [ADD_SPAM_COUNT]

Replace this image path with your actual chart at docs/spam_pie_chart.png

• Training notebook: Spam Email Detection Machine Learning Train Script.ipynb

• Saved models:

  • models/spam_model.pkl
  • models/tfidf_vectorizer.pkl

{
  "text": "Congratulations! You've won a free iPhone. Click here to claim."
}

{
  "prediction": "spam"
}

• notebooks/Api Example.ipynb: Example API interactions
• notebooks/Spam Email Detection Machine Learning Train Script.ipynb: Model training and evaluation

.
├── app.py
├── requirements.txt
├── requirements.lock.txt
├── Dockerfile
├── docker-compose.yml
├── AUTHOR
├── data/
│   └── email.csv
├── models/
│   ├── spam_model.pkl
│   └── tfidf_vectorizer.pkl
├── notebooks/
│   ├── Api Example.ipynb
│   └── Spam Email Detection Machine Learning Train Script.ipynb
└── docs/
    └── spam_pie_chart.png
    └── avatar.png

GNU Affero General Public License v3.0