How Big Is A Dragon?

2025-11-17 · Source: The Ontologist · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Advanced, medium

Summary

The article introduces the Annotation Context Ontology (ACO), a framework designed to manage contextual information within knowledge graphs, moving beyond traditional databases that force a single "version of truth." ACO addresses the challenge of assertions that vary based on factors like source, time, location, or specific conditions, using a dragon length example to illustrate how an entity's properties can change with its "state." It defines four specialized context types: Temporal Context for time-based scoping, Spatial Context for location, Epistemic Context for knowledge source and confidence, and Modal Context for hypothetical scenarios. The framework, available as a first draft on GitHub, allows for the reification of statements to include conditions under which an assertion is considered valid, effectively turning a knowledge graph into a state machine capable of handling dynamic variables and conflicting information from diverse sources.

Key takeaway

For AI Scientists and Research Scientists building knowledge graphs, you should adopt contextual ontologies like ACO to manage the inherent variability and conflicting nature of real-world data. This approach allows your systems to preserve multiple valid perspectives, track data provenance, and model state changes, leading to more nuanced analysis and robust decision-making. Consider how your current data models handle assertions that are true only under specific conditions or from particular sources.

Key insights

Knowledge graphs should manage assertions with context, not just facts, to reflect varying truths.

Principles

• Validity is not reality.
• Provenance builds trust in assertions.
• Context anchors statements to a frame of reference.

Method

The Annotation Context Ontology (ACO) uses reification and specialized context types (Temporal, Spatial, Epistemic, Modal) to define conditional validity for assertions, transforming a knowledge graph into a state machine.

In practice

• Track historical records with conflicting accounts.
• Manage scientific data under varying conditions.
• Integrate multi-source enterprise knowledge graphs.

Topics

• Knowledge Graphs
• Annotation Context Ontology
• RDF-Star
• Contextual Reasoning
• Data Provenance

Code references

• kurtcagle/annotation-context-ontology

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

Related on AIssential

• What Makes Knowledge Graphs AI-Ready — Knowledge Graphs for AI bridge siloed data semantics with organizational ontology for machine-readable reasoning.
• Provenance-Enhanced Statements in Knowledge Graphs — DEC interprets knowledge graph provenance as epistemic stance, enabling principled reasoning over attributed claims and disagreements without collapsing into inconsistencies.
• Context Graphs: A Series of Unfortunate Events — Context graphs model information as a series of time-bound events, contrasting with topical knowledge graphs.
• The Living Graph: Holons and the Four-Graph Model — Holons provide a structured, multi-layered approach to manage complex, hierarchical knowledge graphs with clear boundaries and contexts.
• Ontology, Part I — Conflicting definitions and terminology create significant barriers to understanding and implementing ontologies in AI.
• Why Early Knowledge Graph Adopters Will Win the AI Race — Early adoption of knowledge graphs and ontologies provides a competitive edge in enterprise AI.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Ontologist.