📩 Spam Detection using Google Word2Vec + XGBoost

This project is a spam classification web app built using:

• Google Pretrained Word2Vec (300d)
• Average Word Embeddings
• XGBoost Classifier
• Streamlit Web Interface

---

🚀 Features

• Semantic word embeddings using pretrained Word2Vec
• XGBoost model for classification
• Clean modular architecture
• Confidence score display
• Streamlit-based UI

---

🧠 Model Pipeline

1. Text preprocessing (tokenization, stopword removal, lemmatization)
2. Convert words → Average Word2Vec embedding
3. XGBoost prediction
4. Spam / Ham output

---

📈 Model Performance

Accuracy: 96%

Precision: 98%

Recall: 98%

F1 Score: 98%

---

📂 Project Structure

├── SMS Spam Classification.ipynb ├── SMSSpamCollection.txt
├── app.py
├── embeddings.py
├── google_w2v_model.pkl
├── model.py
├── preprocessing.py
└── requirements.txt

---

⚙️ Setup Instructions

1️⃣ Clone the repository

git clone https://github.com/Master-45-vic/SMS-Spam-Detection-.git
cd SMS-Spam-Detection

2️⃣ Install dependencies

pip install -r requirements.txt

3️⃣ Download Google Word2Vec model

Download the pretrained GoogleNews Word2Vec model separately and place it in the project root folder:

GoogleNews-vectors-negative300.bin

(Note: File is ~1.5GB and not included in repository)

4️⃣ Run the application

streamlit run app.py

📊 Dataset

SMS Spam Collection Dataset from Kaggle.

👨‍💻 Author

Prashanth M

⭐ If you like this project

Give it a star on GitHub!