🛡️ Phishing Website Detection Using Machine Learning

A machine learning-powered web application that detects whether a given website URL is phishing or legitimate using multiple feature types — including URL structure, domain metadata, and HTML/JavaScript behavior.

⚠️ Over 90% of cyber attacks start with phishing. This tool helps users identify potentially harmful URLs before they click.

---

🔍 Features

• ✅ URL-based features: Length, depth, special characters, HTTPS usage, etc.
• ✅ Domain-based features: WHOIS registration check, domain age, DNS availability
• ✅ HTML/JS-based features: Presence of <iframe>, alert() scripts, and suspicious JS functions
• ✅ Binary classification model (Phishing vs Legitimate)
• ✅ Flask-based GUI with modern dark theme
• ✅ Real-time prediction on user input

---

🧠 Models

• Algorithm: RandomForestClassifier

• Training:

  • 10,000 phishing URLs
  • 10,000 legitimate URLs

• Evaluation:

  • Accuracy: ~91%
  • Precision: ~92%
  • Recall: ~88%

• Algorithm: Logistic Regression

• Training:

  • 10,000 phishing URLs
  • 10,000 legitimate URLs

• Evaluation:

  • Accuracy: ~82.1%
  • Precision: ~87%
  • Recall: ~82%

• Algorithm: XGBoost

• Training:

  • 10,000 phishing URLs
  • 10,000 legitimate URLs

• Evaluation:

  • Accuracy: ~82.03%
  • Precision: ~87%
  • Recall: ~82%

• Algorithm: MLP Classifier

• Training:

  • 10,000 phishing URLs
  • 10,000 legitimate URLs

• Evaluation:

  • Accuracy: ~96%
  • Precision: ~96%
  • Recall: ~96%

---

🌐 Live Demo: Click Here

---

⚙️ Tech Stack

Python

Flask

Scikit-learn

BeautifulSoup

WHOIS

💻 Run Locally

🔧 1. Clone the Repository

git clone https://github.com/rahulptl165/Phishing-Website-Detection-Using-ML.git
cd Phishing-Website-Detection-Using-ML

pip install -r requirements.txt

cd app

python app.py

✍️ Authors

Rahul Kumar                 Ayushmaan

💼 LinkedIn                 💼 LinkedIn

💻 GitHub                   💻 GitHub