Phishing Url Detection Using Distlbert And Capsule Neural Networks

Atla, Saketh Reddy (2024) Phishing Url Detection Using Distlbert And Capsule Neural Networks. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (895kB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

This paper presents a new approach to detecting phishing URLs (Uniform Resource Locator) by combining DistilBERT and Capsule Neural Networks. The paper presents a continuous solution for a problem in an ever-altering cyber threat environment. In this study, three models, namely, DistilBERT, Capsule Neural Network, and a new Hybrid of the two, are presented and evaluated using 277,645 real-world 277,645 URLs. The DistilBERT model showed accuracy, precision, recall, and F1-score of 98%, demonstrating its great potential in real-time phishing detection. Quite surprisingly, the Capsule Network showed very poor accuracy of only 50% in these tasks, hence showing large challenges when it comes to the adaptation of this architecture for URL classification. While drastic improvements were noted over DistilBERT, Hybrid showed marginal improvements, specifically for recall alone (98.92% vs. 98%), hence proving it to show potential further in sophisticated manners of phishing detection.

The paper contributes to the literature in general on the efficacy of transformer-based models for URL classification and raises some issues connected with the application of Capsule Networks in security-related, plain-text tasks. The results have huge implications for the development of more sustainable, real-time phishing detection systems and provide a targeting ground for future studies in adaptive learning frameworks, explainable AI, and multimodal approaches toward phishing detection.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Menghwar, Teerath Kumar UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	07 Aug 2025 09:01
Last Modified:	07 Aug 2025 09:01
URI:	https://norma.ncirl.ie/id/eprint/8458

Actions (login required)

View Item