Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{178618,
        author = {Anjali Singh and Prerana Upadhyay and Saumya Yadav and Prakash Khelage},
        title = {Graphsafe: Intelligent Intrusion Detection System with Explainable AI},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {11},
        number = {12},
        pages = {5650-5661},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=178618},
        abstract = {Contemporary cyber threats leverage the relational complexity of networks, frequently bypassing traditional Intrusion Detection Systems (IDS) that process traffic data independently. This thesis introduces a Graph Neural Network-based Intrusion Detection System to counter these shortcomings by representing network entities and their interactions as graphs. Employing Graph Neural Networks (GNNs), GNN-IDS captures the static structure and temporal evolution of network activity, providing a broader perspective than standard techniques. The system combines historical attack graphs with real-time traffic, utilizing Graph Convolutional Networks (GCN), Graph Attention Networks (GAT), and an edge-weighted GCN variant (GCN- EW) to detect intricate attack signatures. It further enhances resilience against adversarial interference and offers interpretable results through uncertainty quantification and attention-driven analysis. Evaluated using synthetic datasets and the CICIDS- 2017 benchmark, GNN-IDS achieves notable accuracy, robustness against tampering, and clarity in tracing intrusion routes. These outcomes demonstrate the power of GNNs to strengthen intrusion detection by exploiting network connectivity, laying groundwork for more effective and transparent cybersecurity solutions. This research elevates IDS performance and invites exploration into graph-centric security approaches.},
        keywords = {Intrusion Detection System (IDS), Graph Neural Networks (GNNs), Network Security, Cybersecurity, Attack Graphs, Anomaly Detection, Adversarial Resilience, Explainability, Network Traffic Analysis, Deep Learning.},
        month = {May},
        }