Detection of Phishing Website Using Gradient Boosting Algorithm

Computer Science and Engineering

Phishing attacks have emerged as a significant risk to online security, frequently deceiving users by impersonating trustworthy websites to obtain sensitive data. Traditional detection methods often struggle to keep up with the ever-evolving tactics employed by phishers. This paper proposes a fast and effective machine learning-based method for identifying phishing websites, using only the URLs as input features. The suggested model identifies and examines a diverse set of lexical and statistical features from URLs, including domain length, the presence of special characters, and specific token patterns, all of which are indicative of phishing activities. Gradient boosting, renowned for its exceptional predictive accuracy and capacity to handle intricate datasets, is utilized to construct a reliable classifier. The model is trained and tested using publicly available phishing datasets and performs exceptionally well, achieving high accuracy, precision, and recall compared to other traditional classifiers. The result showcase the effectiveness of url feature-based detection in identifying phishing websites, providing a lightweight solution for improving cybersecurity measures.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.