CyberHull framework uses convex hulls for deterministic malicious URL detection

Researchers have developed CyberHull, a novel deterministic framework for identifying malicious URLs. This system utilizes convex hull boundaries across various feature domains, including URL structure, hyperlink topology, and semantic content. By employing the QuickHull algorithm, CyberHull constructs compact 'vaults' and classifies URLs through efficient point-in-polyhedron tests. Experiments show CyberHull achieves 92% detection accuracy, significantly outperforming a PCA-based baseline and offering a more computationally efficient alternative to deep learning methods. AI

IMPACT This deterministic approach to URL detection could enhance cybersecurity defenses against sophisticated phishing attacks.

RANK_REASON The cluster describes a new academic paper detailing a novel technical framework. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• arXiv
• CyberHull
• Hugging Face
• principal component analysis
• QuickHull

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →