Phishing URL Checker

A machine learning-based tool to detect phishing and malicious URLs.

Description

This project uses deep learning to classify URLs as either safe or malicious. It employs a convolutional neural network model trained on a large dataset of labeled URLs to identify potential phishing attempts and malware distribution sites.

Features

• URL safety classification using deep learning
• Real-time URL checking functionality
• Simple GUI interface for URL input
• Command line interface for batch processing
• Supports both single URL and text with multiple URLs
• Returns probability scores for malicious classification

Technologies Used

• Python
• TensorFlow/Keras
• Tkinter (for GUI)
• NumPy
• Pandas

Model Details

• Input: URLs encoded as sequences of characters
• Architecture: Convolutional Neural Network with 3 convolutional layers
• Output: Binary classification (Safe/Malicious) with probability score
• Training accuracy: ~95%
• Validation accuracy: ~94%

Usage

NoteBook

python URL_Checker.ipynb

Enter a URL in the input field and click "Check URL" to get the classification result.

Command Line Interface

python tool.py

Input text containing URLs when prompted. The script will extract and check all URLs found.

Installation

1. Clone this repository
2. Install required packages:

pip install tensorflow numpy pandas string

3. Download the pre-trained model file (model_40.keras)
4. Run either the GUI or command line interface

Project Structure

• URL_Checker.ipynb - Main notebook
• test2.py - Command line interface
• models/ - Directory containing trained model
• dataset/ - Training data (CSV format)

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

Dataset and base architecture inspired by various phishing detection research papers and implementations.