How banashree19/real-time-phishing-website-detection works
This is a desktop phishing-URL checker: a user submits a URL, the GUI asks the application layer to extract URL features, then asks a classifier to label the URL and displays the result. Feature extraction also performs reputation lookups and can inspect page content, though page-content features are currently disabled in the sampled code. The GUI uses a Random Forest; other classifier implementations are present but are not selected by that GUI path. The source excerpts show the call structure, but not the contents or exact roles of the CSV files in the tree.
Main entry points: Feature_extraction.py Feature extraction · gui.py URL checker GUI · main.py URL processing · trainer.py Training and prediction
Components 12
User interface
- URL checker GUI
gui.py
Detection workflow
- URL processing
main.py - Training and prediction
trainer.py - Random Forest
trainer.py
Feature extraction
- Feature extraction
Feature_extraction.py - URL signals
Feature_extraction.py - Reputation lookups
Feature_extraction.py - Page content signals
Feature_extraction.py
Other components
- User
- Alexa data service
- Safe Browsing service
- GeoIP ASN data
Connections 11
- User → URL checker GUI: submits URL
README.md - URL checker GUI → URL processing: processes URL
gui.py - URL checker GUI → Training and prediction: requests prediction
gui.py - URL processing → Feature extraction: extracts features
main.py - Feature extraction → URL signals: derives signals
Feature_extraction.py - Feature extraction → Reputation lookups: queries reputation
Feature_extraction.py - Feature extraction → Page content signals: can inspect
Feature_extraction.py - Reputation lookups → Alexa data service: queries ranks
Feature_extraction.py - Reputation lookups → Safe Browsing service: checks URL
Feature_extraction.py - Reputation lookups → GeoIP ASN data: looks up ASN
Feature_extraction.py - Training and prediction → Random Forest: fits and predicts
trainer.py