psychologyPattern — Load → Reason → Query

Materialize the implicit consequences of a graph, then query the closure — so a query sees entailed facts, not just asserted ones.

When to use it

You need queries to return facts that follow from the ontology — subclass membership, transitive relations, inverses — not only the triples literally in the file. The amber Reason step is single-threaded, so the graph must fit your box (if it does not, use the carve pattern).

A worked scenario — an inverse property

ex:owns has an inverse, ex:ownedBy. We assert only that Alice owns a book — never the inverse — and let the reasoner derive it.

Step 1 — load the data

SELECT pgrdf.add_graph(104900);
SELECT pgrdf.parse_turtle('
@prefix ex:  <http://example.org/> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .

ex:owns owl:inverseOf ex:ownedBy .
ex:alice ex:owns ex:book .
', 104900);

Step 2 — query before reasoning

Ask what Alice owns by the inverse — nothing is asserted yet:

SELECT * FROM pgrdf.sparql(
  'PREFIX ex: <http://example.org/>
   SELECT ?s WHERE { ?s ex:ownedBy ex:alice }');
--  → (no rows)

Step 3 — reason

SELECT pgrdf.materialize(104900, 'owl-rl');
--  → {"base_triples": 2, "inferred_triples_written": ..., ...}

The OWL 2 RL prp-inv rule entails ex:book ex:ownedBy ex:alice from ex:alice ex:owns ex:book + ex:owns owl:inverseOf ex:ownedBy.

Step 4 — query after reasoning

Same query, now answered from the closure:

{"s": "http://example.org/book"}

Grounded in the test suite

This is tests/w3c-sparql/49-reasoning-profile-owl-rl. Its sibling 48-reasoning-profile-rdfs shows the contrast: owl:inverseOf is an OWL 2 RL rule, not an RDFS rule — so this same triple would not be entailed under materialize(…, 'rdfs'). The profile you pick changes the closure.

Notes

• materialize is idempotent — re-run it freely; prior inferred rows are dropped and re-derived.
• Pick the profile: 'owl-rl' (default) or 'rdfs' — see Reason.

Next step

Add a conformance gate over the closure with Load → Validate → Query, or right-size a large source with Ingest → Carve → Reason.