A 3-Fold Fused Knowledge Graph Dataset for AI-Assisted Threat Mitigation in Smart Grids (Grid Components, Attacks, Experts)

This dataset presents a novel and comprehensive 3-fold Knowledge Graph (KG) designed as a foundational benchmark for AI-assisted threat mitigation in smart grids. It uniquely fuses disparate ontologies across three critical domains: Power Grid Components (specifically Inverter-Based Resources), Cybersecurity Threats (mapped to MITRE ATT&CK), and human Expert Profiles from academia and industry. The resulting heterogeneous graph explicitly links threats to targeted grid assets and qualified experts to specific mitigation tasks. Constructed through an AI-assisted pipeline, every node is enriched with high-dimensional semantic embeddings generated by Large Language Models (LLMs) from detailed textual descriptions. This dataset provides the essential ground truth structure and rich feature set required for developing and evaluating scalable graph learning models, such as Graph Neural Networks (GNNs), for automated threat analysis and precise expert recommendation. The data is provided in standard JSON format facilitating easy integration into machine learning workflows.