Why is the predicate in a triple considered 'first‑class data'?

Because it is a named resource that can itself be described, typed, and queried.

Because it is the most frequently used part.

Because it is a named resource that can itself be described, typed, and queried.

Because it must be unique across the entire graph.

Because it is immutable once inserted.

If you add the triple `:knows rdfs:domain :Person`, what does that achieve?

It declares schema‑level knowledge: any `:knows` triple must have a subject that is a `:Person` (for validation/reasoning).

It deletes all existing `:knows` triples with subjects that are not `:Person`.

It declares schema‑level knowledge: any `:knows` triple must have a subject that is a `:Person` (for validation/reasoning).

It automatically indexes all `:knows` triples for faster retrieval.

It prohibits adding new `:knows` triples afterwards.

Triples, schemas, and ontologies — Semantic Web Academy

Every fact in a KG is a triple: (subject, predicate, object). This simple form is the only construct you need. For example:

:Alice :knows :Bob .
:Alice :worksAt :Acme .
:Bob :livesIn :Tokyo .

The predicate is a first‑class citizen: it has a URI, and you can attach metadata to it (e.g. :knows rdfs:domain :Person ; rdfs:range :Person .). This makes the data self‑describing and enables reasoning.

A KG can also carry a schema (ontology) as a set of triples that define classes, properties, and their interrelations. For example, :Person rdfs:subClassOf :Agent and :Employee rdfs:subClassOf :Person create a hierarchy. These schema triples live alongside instance triples in the same graph, but they are distinguished by their subject and predicate (e.g., rdfs:subClassOf, rdf:type).

This dual nature — data and metadata in the same model — is what makes KGs machine‑understandable: a client can introspect the graph to learn its shape and infer new facts, without relying on external documentation or hard‑coded knowledge.

Theory