← Back to concept explainers

Concept Fable

Ontology

Conclusion first:Ontology is a structured definition of the concepts, categories, properties, relationships, and constraints in a domain. It helps systems unify "what is what, how they relate to each other, and which inferences are allowed." In enterprise Agents and knowledge graphs, ontologies are used to reduce naming confusion and support relational reasoning, permission boundaries, data integration, and explainable retrieval.

Ontology · Knowledge representation

Read the story first. There is no rush to give a definition here; let the question grow on its own first.

Fable

The apothecary first mapped out the names and roles of all things

More and more pharmacies opened in the city, and the same medicinal herb was called sea-wind vine on East Street, green vine root on West Street, and simply wind root in the accounting office. When patients came in with prescriptions, the herb dispensers often had to guess for quite a while first.

At first the shopkeeper only asked everyone to write labels larger. The labels did become more prominent, but same-name different-things, different-name same-things and contraindication relationships were still mixed together. Once, an apprentice put medicines with similar names together and nearly caused an old man to take the wrong one. Only then did the shopkeeper realize the problem was not unclear writing, but that the shop had no shared set of names and roles.

He invited the senior physician, the herbalist, and the accountant to sit together and first define what constitutes medicinal materials, what constitutes an illness, what constitutes a place of origin, what constitutes a contraindication, and under what circumstances two herbs can be used together. They drew up a registry: every concept had a place, every relationship had a name, every constraint had a source. When a new medicine entered the shop, it first had to be connected to this registry.

Later, when patients used colloquial names, the system could find the proper names; the dispensary could look at inventory and know which medicines could be substituted and which absolutely could not; new apprentices could also trace relationships to find the reason. The name registry did not replace physicians' judgment, but it gave all judgments a common language. Arguments no longer got stuck on "which thing exactly are you talking about?"

The manager invited an old physician to draw a genealogy of the names, aliases, contraindications, substitutes, and broader/narrower relationships in the pharmacy. Every name had to know which medicine it pointed to, and could not be confused with similar medicines.

The apprentice thought the drawing manual was too slow. But the next time a patient came with a prescription from elsewhere, the dispensing master followed the manual to find the same substance under a different name, and also avoided medicines that conflicted with an old illness.

This small incident made everyone stop and review. They did not rush to adopt a louder-sounding narrative; instead they wrote down, one by one, the entry points that failed, the rules that were misunderstood and the boundaries that needed to be kept. Later, when they encountered similar situations, they first looked at the on-site constraints and traces of failure, then decided whether to loosen, tighten or change the approach.

The pharmacy relied less on guesswork from then on. Literal labels remained, but what truly supported dispensing was a set of names and relationships recognized by the whole shop. The shopkeeper said the pharmacy truly became stable only when it began to establish clear names for the world.

Reveal

This story is about: ontology

Ontology is a structured definition of the concepts, categories, properties, relationships, and constraints in a domain. It helps systems unify "what is what, how they relate to each other, and which inferences are allowed." In enterprise Agents and knowledge graphs, ontologies are used to reduce naming confusion and support relational reasoning, permission boundaries, data integration, and explainable retrieval.

What matters is that it is not just a term, but a structural problem that repeatedly appears in real AI / Agent systems. Understanding it lets you judge when to add a model, when to change a process, and when to add governance.

Metaphor mapping

  • Apothecary Name Register: domain ontology
  • Medicinal herbs, symptoms, contraindications: concepts, attributes, and relationships
  • Same name, different thing: semantic conflicts in enterprise data
  • Relying only on shelf labels: surface-level classification is insufficient
  • Compatibility rules: constraints and inferable relationships
  • Fewer wrong medications: unified knowledge structure improves system reliability

Andy Ren Xu's judgment

The practical value of this concept is that it helps you break "AI features that look smart" into deliverable, verifiable, and governable units of work.

schemaknowledge graph
京ICP备2026014272号