Mike's Notes
This is the first of a series of thoughtful opinions about using Ontology in software information systems.
Pipi has an existing BORO Engine (bor) (inspired by Chris Partridge's book "Business Objects: Re-engineering for Re-use"), but it runs in reverse: it imports triples to extract entities and relationships, then uses them to build reference relational databases for back-end industry workspaces.
Part one: This description of ontology is from Kingsley Uyi Idehen, Founder and CEO of OpenLink Software. It's a comment he made about a LinkedIn post from Mansit Suman. I made up the title. Kingsley has a lot of good things to say.
From Google.
" Resource Description Framework (RDF) and implicit Label-based Property Graphs (LPGs) represent two contrasting methods for modeling data and semantics in graphs. RDF relies on formal, machine-readable ontologies (using RDFS/OWL), while LPGs use application-driven, informal key-value labels embedded directly within the data structure.
Core Structural Differences
- RDF Ontologies: Explicitly defined classes and properties with global uniform resource identifiers (URIs).
- LPG Schemas: Implicit, local typing where node labels and edge properties act as ad-hoc tags without a strict global logic layer.
- Standardisation: RDF follows rigid W3C web standards; LPGs prioritise engineering flexibility and speed.
- Query Languages: RDF uses SPARQL; LPGs use platforms like Cypher or Gremlin
" - Google Search AI Mode
Resources
References
- Business Objects: Re-engineering for Re-use, Chris Partridge. 2005.
Repository
- Home > Ajabbi Research > Library >
- Home > Handbook >
Last Updated
14/08/2026
The ontology conversation
Founder & CEO at OpenLink Software | Driving GenAI-Based AI Agents | Harmonizing Disparate Data Spaces (Databases, Knowledge Bases/Graphs, and File System Documents).
The ontology conversation is simply about a machine-computable framework, grounded in logic, for expressing entity and relationship types. That’s it.
An ontology is represented using a variety of document types to describe the nature of entities and relationships. It just isn’t as complicated as ever-confusing marketing communications and opinions would have everyone believe.
The phrase Knowledge Graph is a mainstream marketing colloquialism for entity relationships informed by explicit (in the case of RDF) or implicit (in the case of LPGs) ontologies.
Semantic Layers are fundamentally Knowledge Graphs, but there’s a school of marketing communications that strangely believes they apply uniquely to business-intelligence-oriented metrics, which really doesn’t hold much water under basic scrutiny.
Context Layers are Knowledge Graphs too. They provide session-level frameworks (or harnesses) that enable more deterministic outcomes from AI agents powered by Large Language Models (LLMs).
Most importantly, these entity relationships are best represented using open standards that include RDF (Resource Description Framework). Do so using Linked Data principles, and you end up with a powerful Semantic Web.
No comments:
Post a Comment