The EMERALD project has advanced the development of AIOX, an AI-assisted ontology extension tool designed to support the evolution of the CertGraph ontology. AIOX helps domain experts and ontology engineers identify missing concepts and properties directly from YAML-based security metric definitions and guides their integration into the ontology. By combining semantic analysis with interactive expert validation, the tool reduces the manual effort required to maintain and extend complex cybersecurity knowledge models while ensuring that all changes remain under human control.

Technically, AIOX is implemented as a modular Python-based command-line application operating directly on CertGraph OWL files and security metrics definition YAML files. The tool analyses metric definitions, detects ontology gaps, and applies AI-supported recommendation techniques to suggest parent classes, relations, and data property assignments. Its recommendation engine combines sentence-transformer embeddings with masked language model approaches to improve semantic placement accuracy. Initial evaluations across multiple ontologies demonstrate that metric-derived context provides valuable signals for ontology refinement and can significantly reduce the cost and complexity of ontology engineering in evolving cybersecurity environments.