A comprehensive single model machine learning approach for analysis of homoglyph attacks

Akmal, Danial Irfan Akmal (2024) A comprehensive single model machine learning approach for analysis of homoglyph attacks. Project Report. Melaka, Malaysia, Universiti Teknikal Malaysia Melaka. (Submitted)

Text (Full Text) A comprehensive single model machine learning approach for analysis of homoglyph attacks.pdf - Submitted Version Download (12MB)

Abstract

Homoglyph attacks, a subset of phishing attacks, exploit visual similarities between characters to deceive users, leading to significant cybersecurity threats. This project proposes a comprehensive single-model machine learning approach for analysing homoglyph attacks. By leveraging Information Gain for feature selection, the model identifies and prioritizes critical URL characteristics that distinguish malicious from benign links. The selected features are then used to train a robust machine-learning model that can accurately detect homoglyph attacks. This approach not only enhances detection accuracy but also simplifies the model, making it more interpretable and efficient. The project aims to contribute to improved cybersecurity measures by providing a detailed methodology for identifying and mitigating homoglyph phishing attacks.

Actions (login required)

View Item