The paper introduces a novel approach to Intrusion Detection Systems (IDS), called X-CBA. This approach leverages the advantages of Graph Neural Networks (GNNs) to effectively process network traffic data while implementing a new Explainable AI (XAI) methodology. Unlike most GNN-based IDS, this method uses a broader range of traffic data through network flows, including edge attributes, to enhance detection capabilities and adapt to new threats. The approach achieves high accuracy in threat detection and provides clear explanations of its analytical outcomes. The research aims to bridge the current gap and facilitate the broader integration of Machine Learning/Deep Learning technologies in cybersecurity defenses.
Publication date: 2 Feb 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2402.00839
