Most metadata standards used in research information management — Dublin Core, DataCite’s schema, CERIF — describe records as sets of attributes attached to an object: a dataset has a title, a creator, a publication date. CIDOC CRM, formally the ISO 21127 Conceptual Reference Model, takes a different starting point. It models the domain as a network of events — an object was made, by someone, at a place, at a time, using a technique — and lets everything else (people, places, objects, concepts) be defined by the roles they played in those events. For research administrators and digital-humanities teams working with cultural-heritage, archival, or museum-derived research data, that distinction is the reason CIDOC CRM exists as a separate standard rather than a redundant one.
What CIDOC CRM is
CIDOC CRM is a formal ontology — a defined set of classes and properties with explicit logical relationships — for integrating information from disparate cultural-heritage documentation sources: museum collection-management systems, archaeological excavation records, archives, and library catalogs. Its own documentation describes it as “a theoretical and practical tool for information integration in the field of cultural heritage,” built to let institutions map their existing, heterogeneous local data models onto a shared conceptual structure without forcing them to rebuild those systems from scratch. The point is not to replace institutional databases but to sit above them as a lingua franca: once two collections are each mapped to CIDOC CRM, their records can be queried, linked, and reasoned over together, even though the underlying systems know nothing about each other.
Governance and ISO status
CIDOC CRM is developed and maintained by the CIDOC CRM Special Interest Group (SIG), operating under CIDOC — the International Committee for Documentation of the International Council of Museums (ICOM). It was first accepted as an official ISO standard in December 2006, under the number ISO 21127. The standard has been revised since: the current edition is ISO 21127:2023, which supersedes the prior ISO 21127:2014 text. Governance is institutional rather than corporate — participating organizations are public and private cultural-heritage and research bodies contributing through the SIG’s open working process, and the model itself is published as a versioned, freely available specification at cidoc-crm.org rather than sold as a licensed product.
Why an event-based model, specifically
Object-attribute metadata schemas work well when the facts about a record are stable and independent of each other. Cultural-heritage documentation is rarely that simple: a single artifact may have been made by one person, modified by another decades later, excavated at a specific site and stratigraphic layer, attributed to a period on disputed evidence, and later digitized by a third institution — each of these is a distinct event, with its own actors, place, time span, and evidentiary basis, and several of them can be uncertain, contested, or only partially known. Treating “creator” as a flat attribute of the object loses all of that structure. CIDOC CRM’s event-centric design — centered on classes like E5 Event, E7 Activity, and more specific subclasses such as E12 Production or E7 Modification — lets each of those episodes be recorded as its own node with its own participants and provenance, which is also why the model extends naturally into digital provenance, archaeological stratigraphy, and scientific observation, domains covered by its extension family described below.
How CIDOC CRM differs from CERIF and other research-information data models
Within the CASRAI research-information cluster, the closest comparison point is CERIF, the Common European Research Information Format maintained by euroCRIS. Both are formal data models built to make heterogeneous institutional records interoperable, and both define entities, relationships, and time-stamped roles rather than a flat record structure — but they serve different domains and communities. CERIF’s entity set is built around the research-management lifecycle: people, organizations, projects, publications, equipment, and funding. CIDOC CRM’s classes are built around cultural-heritage documentation: physical and conceptual objects, places, periods, and the production, modification, and interpretation events that connect them. A university research office managing grants and outputs works in CERIF/CRIS terms; an archive, museum, or digital-humanities project describing an artifact, manuscript, or excavation context works in CIDOC CRM terms. The two are not competitors — some digital-humanities infrastructure projects map data between both, and CIDOC CRM’s serializations use the same underlying RDF/linked-data foundations as other semantic-web metadata work, including ontologies such as PROV-O for provenance.
The CIDOC CRM extension family
Because a single monolithic ontology cannot cover every specialized documentation need, CIDOC CRM is deliberately modular: a compact “CRMbase” core, plus a family of compatible extensions developed with specific research communities and formally harmonized with the base model. The most established extensions include:
- CRMarchaeo — archaeological excavation, stratigraphy, and site documentation.
- CRMba — archaeological standing buildings.
- CRMdig — provenance of digital and digitized objects, including the steps of a digitization workflow itself.
- CRMgeo — spatiotemporal reference and geometry.
- CRMinf — argumentation and reasoning, i.e. how a scholarly conclusion is justified from evidence.
- CRMsci — scientific observation and measurement, generalized across disciplines.
- CRMsoc — social phenomena and collective activity.
- CRMtex — critical editions and textual scholarship on ancient/historical texts.
- FRBRoo / LRMoo — bibliographic information, aligned with IFLA’s Library Reference Model (LRMoo is the current, IFLA-aligned successor to the earlier FRBRoo).
- PRESSoo — serials and periodicals.
A research group only adopts the extensions relevant to its material — a digitization project cares about CRMdig, an excavation project cares about CRMarchaeo — while everything still resolves back to the shared CRMbase classes, which is what keeps cross-institution querying possible.
Where CIDOC CRM fits in a research data management workflow
For research administrators, CIDOC CRM matters less as something to implement directly and more as something to recognize and accommodate when it appears in a project’s data. If a funded project involves digitized museum collections, archaeological fieldwork, manuscript or archival digitization, or a digital-humanities corpus built by aggregating cultural-heritage sources, the underlying metadata is increasingly likely to be structured against CIDOC CRM rather than a generic schema like Dublin Core. That has practical consequences for a data management plan: describing the metadata standard accurately (citing ISO 21127 and the specific extensions used, not a generic “custom schema” description), planning for RDF/linked-data export if the funder or repository expects it, and budgeting for the domain expertise CIDOC CRM mapping genuinely requires — it is a more demanding standard to apply correctly than a flat attribute schema, precisely because of the event-based structure described above. See CASRAI’s guide on how to choose a metadata schema for a dataset for the general decision framework this standard fits into.
Frequently asked questions
Is CIDOC CRM the same thing as a metadata schema like Dublin Core?
No. Dublin Core is a flat set of attribute fields (title, creator, date, subject). CIDOC CRM is a formal ontology with typed classes, properties, and an event-centric relational structure — closer to a data model than a field list. The two can coexist: a repository might expose simple Dublin Core records for basic discovery while maintaining richer CIDOC CRM-based data underneath for scholarly integration.
Do I need to use CIDOC CRM if my project just has a spreadsheet of museum object records?
Not necessarily. CIDOC CRM’s value is in cross-collection integration and complex event/provenance modeling — for a single, self-contained inventory, a simpler schema may be entirely adequate. It becomes relevant once data needs to be linked with other institutions’ collections, published as linked open data, or reasoned over programmatically.
Is CIDOC CRM only used by museums?
Its origin and primary community are museum documentation (CIDOC is ICOM’s documentation committee), but its extension family and adoption now span archives, libraries, archaeological fieldwork, and digital-humanities research infrastructure more broadly — anywhere heterogeneous cultural-heritage or humanities data needs to be integrated.
How does CIDOC CRM relate to linked open data and RDF?
CIDOC CRM is commonly serialized as RDF/OWL, making it directly usable in linked-open-data publishing — a CIDOC CRM-mapped dataset can be queried with SPARQL and linked to other RDF resources, which is one reason digital-humanities projects favor it over non-semantic-web schemas when interoperability across institutions is a goal.
