Phishing Detection via Graph Neural Networks Analyzing Domain Infrastructure and WHOIS Attributes

Views: 0
Downloads: 0
DOI https://doi.org/10.55923/aksara.v3i1.3
Hendra Susanto 1 , Dewi Anggraini 2 , Dimas Hidayat 3 , Maya Wahyudi 4 , Reza Wijaya 5
1, 2, 3, 4Department of Computer Science, Institut Teknologi Bandung , 5Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember
1 hendra.susanto@itb.ac.id , 2 dewi.anggraini@itb.ac.id , 3 dimas.hidayat@itb.ac.id , 4 maya.wahyudi@itb.ac.id , 5 reza.wijaya@its.ac.id

Downloads

Adversaries constantly generate algorithmic domain variations to bypass static URL blacklists. We model relationships between registered domains, nameservers, IP subnets, and SSL certificates as a heterogeneous graph, applying Graph Attention Networks to identify malicious clusters with 98.1% precision.
Issue Vol. 3 No. 1 (2024)
Section Articles
Pages 157-170
Published 2024-06-30