SMS Spam Classifier

Project: SMS Spam Classifier — TF-IDF + Multinomial Naive Bayes
Author: AJAYA RAJ A N
Live demo: (Add your Streamlit Cloud / HuggingFace link here after deployment)
Link : https://sms-spam-classifier-my-1st-project.streamlit.app/

Summary

A simple SMS spam classifier that detects whether an SMS is spam or ham (not spam). The project includes:

• Data preprocessing (cleaning, stopwords, stemming)
• TF-IDF feature extraction
• Multinomial Naive Bayes classifier
• Streamlit web app for interactive testing and dataset preview

Folder structure

sms-spam-classifier/
├── app/
│ └── streamlit_app.py
├── data/
│ └── spam.csv
├── models/
│ ├── model.joblib
│ └── vectorizer.joblib
├── src/
│ ├── preprocess.py
│ ├── train.py
│ └── predict.py
├── scripts/
├── requirements.txt
└── README.md

Setup (local)

1. Create & activate virtual environment:

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

2. Install dependencies:

pip install -r requirements.txt

3. Train the model (creates \models/model.joblib\ and \models/vectorizer.joblib):

python src/train.py

4. Run the web app:

streamlit run app/streamlit_app.py

Usage

• Open the Streamlit URL printed in the terminal (https://sms-spam-classifier-my-1st-project.streamlit.app/ ).
• Type a message and press Predict to get label + confidence.

Notes on deployment

• Ensure requirements.txt is present
• Make sure the app entry file is: app/streamlit_app.py
• If models are ignored in .gitignore, upload models/ manually to deployment platform

---

Author

AJAYA RAJ A N
LinkedIn: https://www.linkedin.com/in/ajayaraj98
GitHub: https://github.com/Ajay-araj