Answers are grounded in this paper and cite their sources.
This approach demonstrates improved threat detection accuracy in IoT systems, highlighting the role of knowledge graphs.
This study proposes an intelligent threat identification method based on knowledge graphs to address the challenges of security threat detection, hidden attack chains, and complex feature associations in IoT environments. The approach first extracts key features from multi-source heterogeneous device communication data and constructs a knowledge graph containing devices, protocols, behaviors, and event relationships through semantic modeling to achieve global semantic association representation. A graph embedding mechanism is then introduced to vectorize entities and relationships, while an attention-weighted graph convolution structure is used to fuse and propagate multidimensional features, capturing the global dependencies of potential threat patterns. During the graph reasoning phase, the model enhances the interpretability of abnormal behavior detection through relational aggregation and semantic propagation, and finally employs a classifier to output threat probabilities, completing the entire process from knowledge representation to risk discrimination. Experiments on real IoT security datasets show that the proposed method achieves significantly higher accuracy, recall, precision, and F1-Score than traditional deep learning models. It effectively identifies complex attack behaviors and maintains strong robustness, demonstrating the modeling potential of knowledge graph structures in IoT security and providing a systematic solution for multi-source semantic fusion and intelligent threat detection.
No takes yet. Share an insight, caveat, or question.
Yan et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: