🛡️ PhishGuard — ML Phishing URL Detector

PhishGuard is a machine learning web application that detects phishing URLs in real time. Built with Python, Flask, and scikit-learn, it analyzes 40+ structural features of any URL and classifies it as legitimate or phishing with 95% accuracy.

🔍 What it does

• Analyzes any URL instantly in under 15ms
• Extracts 40+ features like entropy, subdomains, TLD suspiciousness
• Uses an ensemble of Random Forest + Gradient Boosting
• Shows confidence score and human-readable risk breakdown
• Supports batch scanning of up to 20 URLs at once

🚀 Quick Start

1. Install dependencies

pip install -r requirements.txt

2. Run the app

python app.py

3. Open in browser

Visit: http://localhost:5000

---

🧠 ML Features Analyzed

• URL length, entropy, special characters
• Suspicious TLDs (.tk, .xyz, .ml, .cf)
• Brand keywords in subdomains
• IP address as hostname
• HTTPS presence
• URL shortener detection
• Phishing keyword count
• Subdomain depth analysis

🛠️ Tech Stack

• Backend: Python, Flask
• ML: scikit-learn, Random Forest, Gradient Boosting
• Frontend: HTML, CSS, JavaScript
• Libraries: NumPy, Pandas, BeautifulSoup4

🎯 Model Performance

• Test Accuracy: 95%
• Training data: 100 URLs (50 legitimate + 50 phishing)
• Features extracted: 40+
• Analysis time: ~15ms per URL

👨‍💻 Made by

Afradox--