---

📩 Spam SMS Detection ML Project

📌 Project Objective

The goal of this project is to build and deploy a machine learning model that can classify SMS messages as Spam or Ham (Not Spam).
The model is trained using a labeled dataset and deployed for real-world testing.

---

🛠️ Tech Stack

• Python
• Scikit-Learn
• Pandas, Numpy
• Natural Language Processing (NLP)
• NLTK
• Streamlit (for deployment)
• Heroku / Render (optional for web deployment)

---

📚 Dataset

• Dataset Source: Kaggle - SMS Spam Collection Dataset
• Description: 5,500 SMS messages labeled as Spam or Not Spam.

---

📊 Project Stages

1. Data Cleaning
2. Exploratory Data Analysis (EDA)
3. Text Preprocessing (tokenization, stemming, etc.)
4. Model Building (Naive Bayes, Logistic Regression, etc.)
5. Vectorization (TF-IDF, GridSearchCV)
6. Model Evaluation (Accuracy, Precision, Recall, F1 Score)
7. PyCharm App Development (Over Streamlit)
8. Heroku Deployment

---

📊 Model Performance

Metric	Score
Accuracy	97.9%
Precision	97.5%
Recall	96%

---

🚀 Deployment

The SMS Spam Detection model is deployable on Heroku and accessible online!

---

⚙️ Steps to Run the Project

1. Clone the repository:

git clone https://github.com/BleeGleeWee/Spam-SMS-Detection.git
cd Spam-SMS-Detection

2. Install dependencies:

pip install -r requirements.txt

3. Run the Jupyter Notebook:

jupyter notebook spam_sms_detection.ipynb

4. For deployed app:

streamlit run app.py

5. Deploy on Heroku

• Install Heroku CLI
• Run the following:

heroku login
heroku create spam-classifier-app
git push heroku main

• Deployed link✨ Here

---

🌟 FINAL SHOWDOWN:

---

Email/SMS-spam-classifier
│
├── data/
│   └── spam.csv                         # Original dataset (or link to download in README)
│
├── notebooks/
│   ├── 01_data_cleaning.ipynb           # Handling nulls, duplicates, formatting
│   ├── 02_eda.ipynb                     # Visualizations and exploratory analysis
│   ├── 03_text_preprocessing.ipynb      # Tokenization, stemming, stopword removal
│   ├── 04_model_building.ipynb          # Naive Bayes, Logistic Regression, etc.
│   └── 05_model_improvement.ipynb       # TF-IDF, hyperparameter tuning, evaluation
│
├── models/
│   ├── model.pkl                        # Serialized trained model (pickle)
|   └── vectorizer.pkl                   # Trained model then vectorized
│
├── app/
│   ├── app.py                           # App entry point
│   ├── predict.py                       # Handles input, loads model, returns prediction
│   ├── model_loader.py                  # Utility to load the model
│   └── train_model.py                   # Training model before testing   
│                       
│
├── static/
│   └── setup.sh                         # Web Design
│
├── tests/
│   └── test_predict.py                  # Unit tests for prediction logic
│
├── .gitignore                           # Ignore notebooks checkpoints, model files, etc.
├── Procfile                             # For Heroku: e.g., `web: gunicorn app.main:app`
├── requirements.txt                     # All dependencies (Flask/FastAPI, sklearn, etc.)
├── nltk.txt                             # NLTK dependencies (stopwords, punkt)
├── README.md                            # Full documentation 
└── LICENSE                              # MIT or any preferred open-source license

---