Message Predictor

A machine learning–based SMS spam detection system using TF-IDF and Multinomial Naive Bayes, deployed with Streamlit. :contentReference[oaicite:1]{index=1}

🚀 Project Overview

This project implements a text classification model that predicts whether a given SMS message is Spam or Not Spam (Ham). It uses classic Natural Language Processing (NLP) techniques and a Machine Learning model, then provides a simple web interface for real-time predictions.

The model is built using:

• TF-IDF Vectorization for converting text messages to numeric features.
• Multinomial Naive Bayes classifier for detecting spam messages.
• Streamlit for deployment and UI.

🧠 How It Works

1. Text Processing

   • Convert raw SMS text into lowercase.
   • Remove punctuation and stopwords (optional in preprocessing).
   • Convert text into TF-IDF features.

2. Model Prediction

   • A pre-trained Naive Bayes model is used to classify the message.
   • Predictions are shown interactively using Streamlit.

📁 Repository Structure

Message_Predictor/ ├── app.py # Streamlit app code ├── model.pkl # Pickled trained ML model ├── vectorizer.pkl # Pickled TF-IDF vectorizer ├── requirements.txt # Python dependencies ├── setup.sh # Setup script (if applicable) ├── Procfile # For deployment (Heroku/Streamlit etc.) └── .gitignore

🛠️ Installation

1. Clone the repository:

   git clone https://github.com/PARAG-TIWARI/Message_Predictor.git
   cd Message_Predictor

Create a Python virtual environment (recommended):

python3 -m venv venv source venv/bin/activate # macOS / Linux venv\Scripts\activate # Windows Install dependencies:

pip install -r requirements.txt ▶️ Run the App (Locally) Once installed, start the Streamlit app:

streamlit run app.py A browser window will open where you can enter any SMS text to classify it as Spam or Not Spam.

📦 Deployment You can deploy this app to platforms like Streamlit Community Cloud or Heroku using the included Procfile and setup.sh. These files help streamline deployment.

🧪 Model Performance The Multinomial Naive Bayes model with TF-IDF vectorization is a standard and effective approach for text classification tasks such as SMS spam detection, known for high precision and reasonable accuracy on balanced text datasets.

🤝 Contributing Contributions are welcome! Please open an issue or submit a pull request with improvements.

📜 License This project is open source and available under the MIT License.

---

If you want, I can also generate badges (e.g., for license, model accuracy, Python version) or a Project Demo GIF section to make it more engaging! ::contentReference[oaicite:3]{index=3}