A machine learning-based solution designed to detect phishing websites with high accuracy. This project leverages various features and algorithms to identify malicious URLs and protect users from online threats
-
URL Analysis: Examines multiple aspects of URLs to identify phishing patterns
-
Machine Learning Models: Implements various algorithms including:
- Random Forest
- Support Vector Machines (SVM)
- Gradient Boosting
- Neural Networks
-
Feature Extraction: Analyzes URL structure, domain information, and content features
-
Real-time Prediction: Using our Chrome Extension "PhishGuard" which uses the Google Safe Browsing API to warn about malicious websites
- Clone the repository:
git clone https://github.com/arry-codes/phishing-detection.git
cd phishing-detection- Install required dependencies:
pip install -r requirements.txt- Run the file app.py
The model is trained on a comprehensive dataset containing:
- Legitimate websites from Alexa top sites
- Known phishing URLs from PhishTank and other sources
- Various URL and content-based features
Our best performing model achieves:
- Accuracy: 98.7%
- Precision: 97.2%
- Recall: 96.8%
- F1-Score: 97.0%

