Ontologies and Controlled Vocabularies
Controlled vocabularies fix the terms used to describe data; ontologies add the relationships among those terms so machines can reason.
Agreeing on words
When different people describe the same thing with different words, data cannot be combined or searched reliably. A controlled vocabulary fixes an agreed set of terms and their meanings, so that everyone tags data the same way. It removes ambiguity from metadata and makes search and integration work across sources.
From lists to structure
- Controlled vocabulary: an approved list of terms.
- Taxonomy: terms arranged in a hierarchy (broader and narrower).
- Thesaurus: adds synonyms and related-term links.
- Ontology: adds formal relationships and logical rules among concepts.
Ontologies
An ontology defines concepts, their properties, and the relationships between them in a machine-readable way. It can state that one concept is a kind of another, that a property links two concepts, or that certain combinations are inconsistent. This lets software infer facts that were not stated explicitly and check data for logical consistency.
Standards
The semantic web provides standard languages: RDF represents facts as subject-predicate-object triples, and OWL expresses ontologies with formal semantics. Identifiers in these systems are resolvable URIs, so vocabularies can reference and extend one another across the web. This shared machinery is what makes data Interoperable in the FAIR sense.
In scientific data
Using a controlled vocabulary for units, diagnostic types, and physical quantities ensures that a temperature is always labeled and interpreted the same way across datasets, and that a catalog search returns everything relevant. For a fusion program, consistent terminology across simulation outputs, diagnostics, and the published record is what lets datasets from different workstreams be compared and combined without manual reconciliation. See metadata and cataloging.