AI-Based Phishing URL Detection System
ABSTRACT:
Phishing attacks are increasing rapidly due to the growth of online banking, e-commerce, and digital
services. Traditional phishing detection systems rely on rule-based techniques or simple machine
learning models that may not provide high accuracy or transparency. This project proposes an AI-
based phishing URL detection system using a Random Forest classifier.
The system extracts structural features from URLs such as URL length, presence of HTTPS, number of
special characters, and suspicious keywords. These features are used to train a machine learning
model to classify URLs as phishing or legitimate. The system also provides a probability score and
feature importance analysis to improve transparency and user trust. The goal of this project is to
enhance online security by providing a simple, efficient, and user-friendly phishing detection
solution.