In knowledge engineering, an ontology is an explicit specification of the kinds of entities, properties, relationships and constraints used to represent a domain.
Ontologies make domain meaning explicit and interoperable, while taxonomies organize concepts primarily through classification and hierarchy.
activecanonicalv1.0.0
Definition
An ontology specifies what kinds of things exist in a modeled domain and how statements about them can be formed. In formal systems it can define classes, properties, individuals, constraints and inference rules.
Difference from taxonomy
A taxonomy primarily classifies concepts, often through broader and narrower categories. A thesaurus adds preferred labels, synonyms and associative relations. A folksonomy emerges from user-created tags. An ontology can include classification but also expresses richer domain semantics.
Electronic Artefacts position
Electronic Artefacts uses a pragmatic entity and predicate vocabulary. The goal is not maximum formalism. It is enough explicit meaning to validate records, generate pages, publish JSON-LD and preserve relations over time.
Limitations
Ontologies can become expensive, politically narrow or disconnected from use. The model should evolve from real questions and records, with versioning and mappings when terms change.
References
See W3C OWL 2, W3C SKOS, Knowledge Graph, Linked Data and Metadata.