SpamShield is an AI-powered tool that detects spam messages using NLP and machine learning. Enter multiple SMS messages and instantly find out which ones are spam.
- Frontend: Streamlit
- Backend: Python
- ML Model: Multinomial Naive Bayes (Scikit-learn)
- Visualization: Plotly
- Dataset: SMS Spam Collection Dataset
- Predicts whether messages are spam or not
- Shows confidence levels
- Summarizes results with charts
- Fast, offline model (no APIs needed)
git clone https://github.com/aaryanpawar16/SpamShield.git cd SpamShield/backend pip install -r requirements.txt streamlit run app.py
📊 Dataset This project uses the SMS Spam Collection Dataset from UCI/Kaggle.
🏗️ Training the Model cd SpamShield/training python train_model.py This will generate:
spam_model.pkl (the trained model)
vectorizer.pkl (the TF-IDF vectorizer)
💡 Deployment To deploy on Streamlit Cloud:
Push this repo to GitHub
Go to https://streamlit.io/cloud
Link your GitHub repo
Set the main file as backend/app.py
Add environment variable PYTHON_VERSION=3.10
🙌 Credits Dataset: UCI/Kaggle
Built with: Python, Scikit-learn, Streamlit
✨ Demo Check out Demo Video on YouTube https://youtu.be/-zSGI8kVX30?si=ERzUfDQUkTJJyt2w
##Check out project on my Streamlit App Link https://spamshield11.streamlit.app/