Ontology

An ontology is a formal specification of the concepts in a domain, their properties, and the relationships between them — expressed precisely enough that a machine can reason over it, and deliberately independent of any single system that stores the data.

That independence is the distinguishing feature. A data model is built for one warehouse or application; an ontology is meant to be reused by many, defining what a customer is rather than which table holds customers. Its formal axioms — this class is a subtype of that one, these two are disjoint, this property must hold — let software derive facts rather than merely retrieve them. In practice most teams get the essential benefit from a governed semantic layer; formal ontologies earn their extra cost when meaning must be shared across many systems, when relationships are the analytical product, or when machines must reason over the model.

Go deeper: Ontology vs Data Model: Meaning You Can Reuse vs Structure You Can Query