Data product ontology

Research Question

What is the data product ontology, which organisations and communities use it, how is it applied in practice within data mesh and data management architectures, and is it still current relative to competing and complementary standards?

Findings

(Populated from §6 Synthesis above.)

Executive Summary

DPROD is the most explicit public ontology for data products in the consulted evidence, but it has not achieved broad consensus adoption across mainstream metadata platforms.

Its technical model is current because the repository is active in 2026 and the ontology remains published on OMG surfaces, but the publication story is still transitional because the public pages disagree on whether DPROD is a beta, a finalized 1.0 release, or still carrying request-for-comments wording.

DPROD is best viewed as a DCAT-based semantic profile for data products that composes with SHACL, DPV, policy vocabularies, and DPDS concepts rather than replacing those neighboring standards.

For organisations deciding what to adopt, DPROD is relevant and promising for interoperable semantic exchange, but production adoption still requires mapping work into platform-specific models such as OpenMetadata, Collibra, DataHub, or Apache Atlas.

Key Findings

1. DPROD is the only dedicated public RDF and OWL ontology for data products found in the consulted evidence, while the other reviewed artifacts are complementary catalog, privacy, interoperability, or descriptor standards rather than competing ontologies.
2. The published DPROD model is intentionally narrow and reuses DCAT classes for resources, datasets, distributions, and data services, while adding data-product-specific semantics for owner, lifecycle, purpose, domain, ports, datasets, protocol, and security schema type.
3. DPROD remains current in 2026 because its public repository shows active maintenance and the OMG spec index is live, but the standard's publication state is still ambiguous because the official surfaces disagree about whether the release is beta, final 1.0, or still in request-for-comments form.
4. The official DPROD materials use the ontology as a composition layer for real operating concerns, with examples for lineage, data rights, quality, schema, and observability, which indicates that DPROD is meant to work alongside specialized vocabularies instead of encoding every governance concern directly.
5. Public adoption signals exist outside the originating workgroup, including a Zazuko vocabulary entry, a downstream Turtle example repository, and an OpenMetadata ontology alignment, but those signals are still sparse and do not demonstrate broad native ecosystem conformance.
6. Major catalog platforms publicly document their own data-product operating models or extensible metadata systems, which indicates that DPROD is currently more of an interoperation target than the native internal schema of the dominant tools.
7. Adjacent standards collectively provide the main alternative capability set to DPROD in the consulted evidence, with DCAT handling catalog exchange, DPV handling privacy semantics, schema.org supporting web discovery, FAIR stating interoperability requirements, and DPDS capturing broader descriptor structure and contracts.
8. DPROD remains relevant as the most current dedicated semantic candidate for data-product interoperability, but organisations should expect local mapping work because no public evidence in this item shows industry-wide consensus or out-of-the-box support across the leading catalog products.

Assumptions

• Assumption: Public product documentation is a reasonable indicator of whether a catalog platform treats DPROD as a first-class supported standard. Justification: Ontology-native support would normally affect platform entity design, documentation, or API references.

Analysis

The evidence weighs most strongly toward DPROD being the leading dedicated semantic candidate because it is the only reviewed artifact that actually publishes an ontology and shapes for data products, rather than only product documentation or a descriptor template.

The competing interpretation is that mainstream tool adoption currently favors product-local models over shared ontologies, and the consulted DataHub, OpenMetadata, Collibra, and Atlas sources are consistent with that reading.

That rival interpretation does not displace DPROD, because the same evidence also shows those tools expose extension points, data-product concepts, or alignment hooks rather than a competing public ontology with comparable semantic scope.

The strongest challenge to treating DPROD as settled is not technical weakness but publication inconsistency, because the namespace and version story diverges across official surfaces.

That inconsistency lowers confidence in claims about formal status, but it does not overturn the central conclusion that DPROD is current and relevant as an interoperability profile built from the same semantic-web stack described in Mitchell (2026) on production web ontologies.

Risks, Gaps, and Uncertainties

• Public independent case studies naming production DPROD deployments remain sparse.
• The consulted official publication surfaces do not fully agree on canonical namespace and release state.
• OpenMetadata's DPROD-aligned ontology section appears broader than the currently published official ontology file, which creates uncertainty about strict conformance.
• This item does not make a clause-level comparison to ISO 8000 because the consulted public ISO 8000 overview page did not provide enough semantic detail to justify one.

Open Questions

• Will OMG stabilize the canonical namespace and release narrative around DPROD 1.0 in a way that removes the current beta versus final ambiguity?
• Will DPROD remain narrowly aligned to DCAT resource, dataset, and service semantics, or expand toward richer port and contract classes closer to DPDS?
• Which major catalog vendors, if any, will publish native DPROD import, export, or mapping support rather than platform-specific models?
• Would a future DCAT profile or W3C-hosted profile reduce adoption friction more effectively than a standalone OMG-centered publication path?

sources

• [x] Object Management Group (2024) DPROD public comment announcement
• [x] Object Management Group (n.d.) DPROD specification index
• [x] EKGF (n.d.) DPROD specification site
• [x] EKGF (n.d.) Adopt DPROD
• [x] EKGF dprod README
• [x] EKGF dprod ontology on develop branch
• [x] EKGF dprod commit ed07bf2 (2026-05-07)
• [x] EKGF dprod commit bf2eec0 (2026-05-07)
• [x] Object Management Group (2024) DPROD ontology TTL
• [x] Object Management Group (2024) DPROD shapes TTL
• [x] W3C (2024) Data Catalog Vocabulary (DCAT) Version 3
• [x] W3C (2024) DCAT 3 Turtle vocabulary
• [x] W3C (2017) Shapes Constraint Language (SHACL)
• [x] W3C (n.d.) Data Privacy Vocabularies and Controls Community Group
• [x] schema.org (n.d.) Dataset
• [x] GO FAIR (n.d.) FAIR Principles
• [x] Open Data Mesh Initiative (n.d.) Data Product Descriptor Specification 1.0.0
• [x] EKGF dprod ontology README
• [x] EKGF dprod examples README
• [x] EKGF dprod concept README
• [x] DataHub (n.d.) Metadata Standards
• [x] DataHub (n.d.) Extending the metadata model
• [x] OpenMetadata (n.d.) Domains and Data Products overview
• [x] OpenMetadata (n.d.) Adding Data Products
• [x] OpenMetadata (n.d.) Data Products API reference
• [x] OpenMetadata ontology file with DPROD-aligned section
• [x] Collibra (2025) About data products
• [x] Collibra (2025) Configuring and building data products
• [x] Apache Atlas (2020) Type System
• [x] Zazuko rdf-vocabularies DPROD prefix metadata
• [x] hyprcat example using DPROD prefix
• [x] ISO (n.d.) ISO 8000 overview page
• [x] Mitchell (2026) Web ontologies in production Knowledge Graphs for multi-step Artificial Intelligence agents