Ontology, Part III

2026-03-09 · Source: Intentional Arrangement · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Expert, quick

Summary

This essay details the construction of the custom elements for the NTWF workflow ontology, building upon prior work that established theoretical groundwork and an engineering methodology. It introduces nine custom classes, thirteen object properties, and five datatype properties, each aligned with specific competency questions. The author justifies each custom term by identifying gaps in existing ontology standards, providing a clear design rationale. The article also covers the ABox instance data used to exercise every term, the ontology validation framework confirming 104 logical consequences, and the reasoner's output. A key aspect discussed is the metadata layer's role across all three architectural boxes, serving as a primary integration surface and often misunderstood as merely documentary.

Key takeaway

For AI Scientists developing custom ontologies, ensure every new term is rigorously justified by a gap in existing standards. Prioritize clear, concise definitions for all custom elements before proceeding to axiom creation; difficulty in defining a concept signals a need for further conceptual clarity, preventing modeling errors and ensuring the ontology's logical coherence and utility.

Key insights

Custom ontology terms are justified by standards gaps and defined clearly before axiom creation.

Principles

• Custom terms require standards survey justification.
• Clear definitions precede axiom modeling.

Method

The method involves defining competency questions, applying design heuristics, building CBox vocabularies, then creating custom classes and properties, and finally validating with ABox data and a reasoner.

In practice

• Use competency questions to define ontology scope.
• Write labels/comments before axioms.
• Validate with ABox instance data.

Topics

• Ontology Engineering
• Workflow Ontologies
• Knowledge Representation
• Semantic Web Standards
• Metadata Architecture

Best for: AI Scientist, AI Architect, Data Engineer, Research Scientist

Related on AIssential

• Ontology, Part I — Conflicting definitions and terminology create significant barriers to understanding and implementing ontologies in AI.
• What Do You Need to Create a Useful Ontology? — Building effective ontologies requires clear scope, purpose-driven design, and careful selection of semantic tools.
• Ontology, Part II — Effective ontology design requires structured thinking, competency questions, design heuristics, and a three-box architecture.
• Ontologies Are the Intentional Core of a True Knowledge Graph — Formal ontologies provide the stable, intentional semantic backbone essential for reliable enterprise AI and knowledge graphs.
• Do You Need An Upper Ontology? — Upper ontologies are often unnecessary and can complicate domain modeling, especially with modern AI and knowledge graph tools.
• Ontology Engineering Tradecraft: This Is Ours, Own It — Ontology engineering uniquely focuses on systematic disambiguation to achieve robust interoperability and information quality.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Intentional Arrangement.