🔐 Email Phishing Detection using Machine Learning

🚀 Overview

This project detects phishing emails using machine learning by analyzing textual content.

It uses TF-IDF vectorization and a Naive Bayes classifier to classify emails as phishing or legitimate.

---

🎯 Problem

Phishing emails are a major cybersecurity threat, often used to steal credentials or sensitive information.

Detecting them automatically is critical for improving email security systems.

---

🧠 Approach

1. Dataset

• Source: Kaggle phishing email dataset

• Multiple datasets combined (Enron, SpamAssassin, etc.)

• Text column: text_combined

• Labels:

  • 0 → Legitimate
  • 1 → Phishing

---

2. Preprocessing

• Train-test split (80/20, stratified)
• Text cleaned and vectorized using TF-IDF

---

3. Feature Engineering

• TF-IDF with:

  • Stopword removal
  • Max features: 5000

---

4. Model

• Multinomial Naive Bayes
• Trained on TF-IDF features

---

📊 Results

Accuracy: 96%
Precision (Phishing): 0.98
Recall (Phishing): 0.95
F1-score: 0.96

👉 Strong performance with balanced precision and recall.

---

📊 Example Predictions

Input:
"Urgent! Your bank account has been suspended. Click here to verify your password."

Output:
Prediction: Phishing
Confidence: 0.988

Input:
"Hi team, please find attached the meeting agenda for tomorrow."

Output:
Prediction: Legitimate
Confidence: 0.993

---

🛠️ Tech Stack

• Python
• scikit-learn
• Pandas / NumPy
• TF-IDF Vectorization

---

▶️ How to Run

pip install pandas scikit-learn kagglehub
python main.py

---

📌 Future Improvements

• Try advanced models (Random Forest, XGBoost, Transformers)
• Deploy as API for real-time email scanning
• Build browser/email client integration
• Add explainability (why classified as phishing)

---

🤝 Why This Project Matters

• Demonstrates applied machine learning
• Real-world cybersecurity use case
• End-to-end pipeline: data → model → evaluation

---

📎 Author

Sanju – ML + Cybersecurity focused developer