Phishing URL dataset

Published: 9 January 2026| Version 1 | DOI: 10.17632/3jddhy2f6s.1
Contributor:
Tandin Wangchuk

Description

This is the phishing URL dataset developed for the research titled "A multi-branch hybrid architecture of transformer, CNN, and a deep feed-forward network with MoE for phishing URL detection". It consists of total of 149,726 URLs, of which 94,919 were legitimate, and 54,807 were phishing URLs. The dataset was developed using the Common Crawl (CC-MAIN-2024-33) and Phishtank phishing URL list. The dataset can be used anyone interested in conducting research or developing machine learning based model for phishing detection.

Files

Institutions

  • Jochi Daigaku

Categories

Cybersecurity, Machine Learning, Deep Learning

Licence