A simple and fast Flask-based REST API that classifies incoming text messages as spam or ham (not spam) using a trained Machine Learning model built with scikit-learn.

• RESTful API built with Flask
• Spam detection using Multinomial Naive Bayes and CountVectorizer
• Language restriction (supports only English)
• Model training script included
• Uses .venv for dependency isolation

spam-detection-api/ ├── app/ │ ├── __init__.py # Flask app │ ├── model.py # App routes │ ├── model.py # Trained model function ├── models/ │ └── model.pkl # Saved trained model ├── data/ │ └── spam.csv # Dataset ├── train_model.py # Train the model ├── run.py # Runs the Flask App ├── requirements.txt # Project dependencies └── README.md # This file 

You can use Docker to build and run this project or follow the steps to setup the project on your machine.

git clone https://github.com/supr1yo/spam-detection-api.git cd spam-detection-api

python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate

pip install -r requirements.txt

Download the spam.csv dataset (e.g., from Kaggle) and place it in the data directory if missing.

python train_model.py

This script reads spam.csv, trains the model, and saves model.pkl inside the models/ directory.

python run.py

The API will start at http://127.0.0.1:5000/

Classifies a message as spam or not.

{ "message": "Free $50 steam points" }

{ "label": "spam", "message": "Free $50 steam points" }

Returns 400 Bad Request if:

• Message is missing from request body

• Algorithm: Multinomial Naive Bayes
• Feature Extraction: CountVectorizer

MIT License. See LICENSE file for details.

Built with 💙 by supr1yo