Linked Data Serialisation¶

Source version 1.2.0. Draft for discussion and validation.

9. Linked Data Serialisation Overlay¶

9.1 Purpose¶

The Linked Data overlay defines how a profile-conformant GS1 artefact is represented as an RDF graph and connected to other GS1 and external resources.

9.2 Required representations¶

The selected representation set should be based on the artefact and intended consumers. A typical package may include:

• JSON-LD for Web and developer integration;
• Turtle for ontology and expert review;
• RDF dataset formats where named graphs or provenance bundles are required;
• HTML with embedded structured data for human and machine discovery;
• SHACL shapes in Turtle or JSON-LD;
• a JSON-LD context where compact developer-facing JSON is needed.

9.3 Linked Data rules¶

1. Use canonical HTTPS IRIs for addressable GS1 resources where an approved identifier policy exists.
2. Distinguish the conceptual artefact from a version and from each distribution.
3. Make IRIs dereferenceable to useful metadata or representations.
4. Support content negotiation where operationally practical.
5. Use explicit RDF types.
6. Reuse established vocabularies before defining new terms.
7. Publish the meaning and governance of every GS1-specific term.
8. Qualify mappings and references rather than relying indiscriminately on owl:sameAs.
9. Represent provenance at artefact and distribution level.
10. Publish SHACL constraints for expected graph structures.
11. Preserve language tags and datatype semantics.
12. Provide deterministic identifiers or skolem IRIs where blank nodes would impede citation or comparison.

9.4 Candidate vocabulary stack¶

9.5 Linked Data package example¶

publication-package/
├── manifest.jsonld
├── metadata.ttl
├── artefact.jsonld
├── artefact.ttl
├── shapes.ttl
├── context.jsonld
├── provenance.ttl
├── examples/
│   ├── valid-example.jsonld
│   └── invalid-example.jsonld
└── validation-report.ttl