What serialization of RDF should I publish?

Priority 4, Best Practice Established

Problem Statement:

RDF is an abstract graph model and there are multiple valid serialization of RDF that can be published as part of a data dump. Common ones include Turtle, JSON-LD, and RDF-XML.

For a consistent user experience, it is beneficial to choose one or more consistent serializations to publish for a data dump.

Best Practice:

For the AAC project, the primary serialization will be N-Triples.

Additionally, the Getty Museum is working on a set of contexts and frames for JSON-LD. Currently generating an AAC context and frame is out of scope, but we will evaluate the Getty's version and see if it is applicable to the project.

Additional serializations may be made available as resources permit.

Discussion:

Turtle vs JSON-LD?

(From Vladimir)

These are merely different RDF serializations: a proper repo can return either.

  • Turtle is more readable, so we should use it for discussions and examples
  • JSONLD can be more easily consumable by your app, especially if we can make proper Framing based on the "business object graph" shape.

(From Rob)

JSON-LD vs anything else? JSON-LD.

Linked Data Fragments

Are LOD Fragments as a possible technology for use instead of SPARQL?

(From Rob)

LD Fragments for CRM would be at best mediocre IMO, but would be happy to see a demonstration of their value in practice.

(From Vladimir)

They're not at all incompatible. LDF implements SPARQL in the browser, by asking only simple patters out of the server (<s p o> where each position can be unbound). LDF is ok for fetching objects, but not ok for finding complex patterns, especially where filters are involved.

(From David)

Are we generating a JSON-LD framing for CIDOC? Does one exist we can steal?

(From Rob, via email, 9/19/2016)

JSON-LD, Though the context is a pain, as there's no good way to handle the predicates with the same human readable name and different URIs :(

(From David, via email, 9/19/2016)

I'm fine with that, but if it's actually going to be /nice/ JSON-LD, then someone needs to write a very nice context, as well as a frame for each institution's model (or one for the super-model of all the institutions, which doesn't exist, despite Vlad and I wanting one).

Reference:

  • TODO: Add links to definitions of the various formats.

Linked Open Data FAQs

Defining Types

How do I specify types for entities?
Priority 1
✔ Best Practice Established
How do I specify types for predicates?
Priority 1
✔ Best Practice Established
What existing extensions to the CIDOC-CRM should I use?
Priority 2
✔ Best Practice Established

Defining URL Structures

What URL should I use for unknown Actors?
Priority 2
✔ Best Practice Established
What is the root URL for each AAC Partner?
Priority 2
✔ Best Practice Established
What is returned when a URL is dereferenced?
Priority 3
✔ Best Practice Established
Which ID is most appropriate for URL construction?
Priority 3
✔ Best Practice Established

Labeling

What are best practices for modeling text strings?
Priority 2
✔ Best Practice Established
What is best practice for labeling external authorities?
Priority 3
✔ Best Practice Established
How do I handle strings in languages other than english?
Priority 4
✔ Best Practice Established

Modeling

How do I handle complexity in knowledge representation?
Priority 1
✔ Best Practice Established
How do I model lists of entities or multiple values?
Priority 2
✔ Best Practice Established
How should I model parts of Actor names?
Priority 2
✔ Best Practice Established

Reconciliation

How do I reconcile objects to authorities?
Priority 2
✔ Best Practice Established
Which entity should I link to in an authority file?
Priority 2
✔ Best Practice Established

Triplestores, RDF, and Inferencing

Which namespace should I use for the CIDOC CRM as LOD?
Priority 1
✔ Best Practice Established
How do I create an RDF representation of an entity?
Priority 3
✔ Best Practice Established
Where should AAC-created vocabularies be hosted?
Priority 3
✔ Best Practice Established
What serialization of RDF should I publish?
Priority 4
✔ Best Practice Established