New research predicts multi-vulnerability attack chains using SBOM graphs

Researchers have developed a novel approach to predict multi-vulnerability attack chains within software supply chains by leveraging Software Bill of Materials (SBOM) data. This method treats SBOMs as dependency-constrained evidence graphs, incorporating component and vulnerability information. A Heterogeneous Graph Attention Network (HGAT) was trained to classify components associated with vulnerabilities, achieving 91.03% accuracy and a 74.02% F1-score. Additionally, a Multi-Layer Perceptron (MLP) model was used to predict cascading vulnerabilities, demonstrating a ROC-AUC of 0.93 on documented attack chains. AI

IMPACT Introduces a novel graph-learning approach for predicting complex software supply chain attacks.

RANK_REASON Academic paper detailing a new methodology for predicting software vulnerabilities. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Armin Moin
• CycloneDX
• Heterogeneous Graph Attention Network
• HGAT
• multilayer perceptron
• Wild SBOMs

• paper
• safety

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