COntExt: Towards Context-Aware Ontology Extension f...

COntExt: Towards Context-Aware Ontology Extension f... | AI Research

Key Takeaways

COntExt is a framework designed to automate the extension of domain ontologies by utilizing structured operational metric definitions.
Organizations increasingly define operational metrics in structured, machine-readable formats to monitor systems, processes, and compliance.
These metric definitions implicitly encode domain knowledge, such as referencing concepts, properties, and relationships, that often extends what is captured in formal ontologies.
Yet the connection between operational metric catalogues and ontological knowledge remains manual, ad-hoc, and labor-intensive.
The framework defines the extension problem as three sub-tasks: parent class prediction, relation type prediction, and data property assignment.

Paper AbstractExpand

Organizations increasingly define operational metrics in structured, machine-readable formats to monitor systems, processes, and compliance. These metric definitions implicitly encode domain knowledge, such as referencing concepts, properties, and relationships, that often extends what is captured in formal ontologies. Yet the connection between operational metric catalogues and ontological knowledge remains manual, ad-hoc, and labor-intensive. We present COntExt, a framework for context-aware ontology extension that takes structured metric definitions as input and suggests how referenced concepts and properties should be integrated into an existing ontology, utilizing the context of these metrics. The framework defines the extension problem as three sub-tasks: parent class prediction, relation type prediction, and data property assignment. Across four cybersecurity ontologies, we evaluate different algorithms for each task. Our results show that metric-derived context improves the suggestions over ontology-context baselines for relation type prediction and data property assignment. Our work demonstrates that operational metric catalogues are a practical and underexploited source for ontology extension. This work enables organizations to maintain their ontologies at a significantly lower cost than manual engineering.

COntExt is a framework designed to automate the extension of domain ontologies by utilizing structured operational metric definitions. Organizations often maintain formal ontologies to represent domain knowledge, but these frequently fall behind as new concepts and relationships emerge. While organizations also maintain operational metrics—such as YAML-based cybersecurity rules—this information is typically ignored during ontology maintenance. COntExt bridges this gap by parsing these metric definitions to suggest updates to an existing ontology.

The COntExt Approach

The framework treats ontology extension as three distinct sub-tasks: predicting the parent class for a new concept, determining the type of relationship between two classes, and assigning data properties to the correct owner class. COntExt is algorithm-agnostic, meaning it allows users to plug in different suggestion engines. The framework functions by taking an ontology and a set of metric definitions as input, parsing the metrics for references to concepts, and using the context found in descriptions, comments, and configuration values to inform its suggestions.

Context Enrichment

A core feature of the framework is its ability to perform context enrichment. By layering metric-derived information—such as thresholds, target values, and human-readable comments—onto the input, the system aims to provide more relevant suggestions than those based on ontology structure alone. The authors evaluated this by comparing performance using only ontology-based context against performance using both ontology and metric-derived context.

Evaluation Results

The researchers tested COntExt across seven ontologies, including those focused on finance, IoT, and cybersecurity. The empirical results indicate that incorporating metric-derived context improves suggestion quality for relation type prediction and data property assignment. However, the study found that this additional context can actually degrade accuracy for parent class prediction.

Practical Implications

This framework provides a way for organizations to keep their ontologies aligned with evolving operational requirements at a lower cost than manual engineering. By treating operational metric catalogues as a source of domain knowledge, the authors suggest that organizations can reduce the labor-intensive nature of manual ontology maintenance. The source code for the framework is publicly available for adaptation to similar use cases.