Back to Reference
Contents
Reference
798 words
Jun 25, 2026

Heratio Help Center article. Category: Federation.

CIDOC-CRM Export

Version: 1.0 Date: 2026-05-26 Author: The Archive and Heritage Group (Pty) Ltd

What it does

The CIDOC-CRM export emits a single archival record as a CIDOC Conceptual Reference Model (CRM) 7.1.3 RDF document. CIDOC-CRM is the de-facto interoperability ontology for museum and cultural-heritage data, used by ResearchSpace, the Erlangen-CRM importer, Getty's Linked Open Data services, and most major aggregator pipelines outside the archival sector.

Heratio's native ontology is RiC-O (ICA's Records in Contexts). The CRM export bridges the two so a Heratio record can be ingested anywhere a CIDOC-CRM consumer expects to see data.

Endpoint

GET /admin/export/crm/{slug}
GET /admin/export/crm/id/{id}

Both routes accept the same query parameters:

Parameter Default Notes
culture session locale ISO 639-1 language code (en, af, fr, etc.)
format rdfxml rdfxml, turtle, or ttl

You can also pick the format via the standard HTTP Accept header:

  • Accept: application/rdf+xml -> RDF/XML
  • Accept: text/turtle -> Turtle
  • no header / anything else -> RDF/XML (default)

The ?format= query parameter wins over Accept when both are supplied.

Examples

Fetch the RDF/XML representation of a record by slug:

curl -H 'Accept: application/rdf+xml' \
  https://heratio.example.org/admin/export/crm/my-record-slug

Fetch the Turtle representation of the same record:

curl -H 'Accept: text/turtle' \
  https://heratio.example.org/admin/export/crm/my-record-slug

Same thing using the ?format= override:

curl https://heratio.example.org/admin/export/crm/my-record-slug?format=turtle

Response headers

Header Value
Content-Type application/rdf+xml; charset=utf-8 or text/turtle; charset=utf-8
Content-Disposition inline; filename="cidoc-crm-<id>.<ext>"
X-CRM-Version 7.1.3
X-Bridge-Phase 659.phase-1

What's in the document

For every record the export emits:

  • The CHO node typed as crm:E73_Information_Object, with rdfs:label, crm:P102_has_title, crm:P1_is_identified_by (the archival identifier), crm:P3_has_note (scope and content).
  • One crm:P14_carried_out_by link per creator actor.
  • One crm:P4_has_time-span link per dated event.
  • crm:P129_is_about literals for subject access points.
  • crm:P7_took_place_at literals for place access points.
  • crm:P72_has_language literals for language access points.
  • crm:P50_has_current_keeper link to the holding repository.
  • One Actor node per creator, typed as crm:E21_Person / crm:E40_Legal_Body / crm:E74_Group based on the actor's entity type.
  • One crm:E40_Legal_Body node for the repository.
  • One crm:E52_Time-Span node per dated event, carrying crm:P82a_begin_of_the_begin and crm:P82b_end_of_the_end as xsd:date literals.

When to use it

  • Submitting a record to a CIDOC-CRM aggregator (e.g. an art-museum union catalogue, a ResearchSpace tenant, a Getty pipeline).
  • Cross-walking Heratio holdings into a museum collections system that speaks CRM natively.
  • Sharing a single record with a researcher who wants the data in a CRM-native triple store.

For bulk export to Europeana use the EDM export instead (see "Europeana EDM Publish"). EDM is purpose-built for Europeana's ingest gate; CRM is the broader interoperability format.

Authentication

The endpoint is behind the standard web middleware group. Admin sessions can browse to the URL directly. Headless clients should either reuse the session cookie of an authenticated browser or proxy through the /api/v2/ token-authenticated alias once it ships.

Known gaps

Phase 1 ships the per-record export only. The following work is tracked under issue #659 for later phases:

  • CRM-aware SPARQL endpoint. Today's /api/sparql proxy returns RiC triples only. Phase 2 extends it to accept crm:E73_Information_Object ?p ?o queries.
  • Getty reconciliation. Phase 3 promotes subject + access-point literals into owl:sameAs links onto Getty AAT/ULAN/TGN IRIs.
  • Reciprocal RiC <-> CRM owl:sameAs on the per-record document, so round-trippers can re-attach the original RiC IRI after a CRM consumer enriches the graph.

Troubleshooting

  • 404 "Record not found": the slug does not resolve via the slug table. Confirm the record exists and the slug is current.
  • 404 "Record not found in culture XX": the record has no i18n row for the supplied locale. Either supply ?culture=en or have the record translated.
  • Empty Actor block: no event row with type_id = 111 exists for the record. CRM crm:P14_carried_out_by is only emitted when a creation event ties an actor to the record.
  • Empty Time-Span block: the events on the record have no start_date / end_date / date_display. Add at least one of the three on the event row.
  • "Europeana EDM Publish" - sibling per-record exporter for the Europeana pipeline.
  • "OpenRiC Fuseki Instance Load" - bulk-loads RiC place and agent IRIs into Fuseki for federated querying.
  • "RiC Dashboard User Guide" - Heratio's RiC-O CRUD surface.