Open access
Aug 2026
GRAL: A GNN-RAG-LLM Framework for Intelligent Cybersecurity Alert Correlation and Analysis
This paper proposes GRAL, which is an AI-driven framework that combines graph neural networks (GNN) for cross-asset temporal alert correlation, retrieval-augmented generation (RAG) for dynamic threat intelligence enrichment, and large language models (LLM) for semantic reasoning and verdict generation.
Deng Zhang, Juan Wang, Hanjun Gao et al.
· Symmetry · 0 citations