A simple NLP-based spam classifier that detects whether a message is Spam or Not Spam using a trained machine learning model. The API is built with FastAPI and deployed using Railway.
Text preprocessing (lowercasing, punctuation removal, etc.)
Bow, TF-IDF vectorization
Naïve Bayes/ Logistic Regression classifier
FastAPI backend for predictions
Deployed online for real-time access
Algorithm: Logistic Regression - 98% accuracy
Vectorizer: BoW
Dataset: SMS Spam Collection Dataset
🌐 Hosted on: https://spamdetection-production.up.railway.app
Try sending a POST request to /predict:
IN bash terminal-
curl -X POST https://spamdetection-production.up.railway.app/predict
-H "Content-Type: application/json"
-d '{"text": "Congratulations! You won a free iPhone. Click here to claim."}'
-
Clone the repo: git clone https://github.com/yourusername/spamdetection.git cd spamdetection
-
Create a virtual environment and install dependencies: python -m venv venv source venv/bin/activate # or
venv\Scripts\activateon Windows pip install -r requirements.txt
Run the API: uvicorn app:app --reload
spamdetection/
│
├── app.py # FastAPI app
├── models/logreg_bow.pkl # Trained Logistic Regression model
├── vectorizer/bow_vectorizer.pkl # TF-IDF vectorizer
├── test/response_test.py # Pre-deifned test phrase
├── requirements.txt # Python dependencies
└── README.md # Project documentation
You can integrate this with the Gmail API to automatically detect spam from your inbox and log the results. (Coming soon...)
MIT License