Changes Coming to Getty Vocabularies LOD

[Getty Vocabularies LOD]

What’s Happening

Coming soon, the Vocab LOD site (vocab.getty.edu) will be changing how it does content negotiation for specific serializations in order to offer 2 modeling options for consumers of Getty Vocabularies URIs: the original Vocab LOD SKOS/Schema.org model and the new Linked.Art modeled Vocabularies data for JSON and JSON-LD. 

This means that if you ask for JSON or JSON LD from the vocab.getty.edu domain by specifying the proper extension or by using the “Accept” header, you will get Linked.Art JSON. Backwards compatibility for the SKOS/Schema.org format will be available by including the proper “Accept-Profile” header.

When the Change is Happening

August, 2022. Stay tuned to this forum and the Vocabs LOD home page for updates.

Background

The original LOD model launched in 2015 used a combination of the SKOS and Schema.org specifications and is currently the only format provided at vocab.getty.edu. In 2021, the Getty Vocabularies was integrated into the overall Getty LOD architecture at data.getty.edu using the Linked.Art (https://linked.art/) model. The activity stream containing all create, update, and delete operations of Vocabularies records for the Linked.Art version of the Vocabularies LOD can be found at:

https://data.getty.edu/vocab/activity-stream

The SPARQL endpoint can be found at:

https://data.getty.edu/vocab/sparql

An additional SPARQL UI for this endpoint is provided at:

https://data.getty.edu/vocab/sparql-ui

The next step in this process is to offer both the SKOS/Schema.org version and the Linked.Art version of the Vocabularies LOD at vocab.getty.edu using the different serializations that present the best compatibility with the model. Linked.Art along with its context document (https://linked.art/ns/v1/linked-art.json) works primarily for a JSON-LD serialization that allows users to expand a JSON structure into RDF triples. For this reason, the JSON and JSON-LD serializations at vocab.getty.edu will conform to the Linked.Art model for TOP LEVEL vocabularies entities (AAT, TGN, ULAN). All other serializations will provide the same SKOS-based modeling the site provides now.

Updated User Guide

Linked.Art Modeled Semantic Resolution (AAT example):

http://vocab.getty.edu/aat/300198841.json

http://vocab.getty.edu/aat/300198841.jsonld

HTTP request of base entity endpoint (e.g., http://vocab.getty.edu/aat/300198841) with “Accept” headers for these serializations:

“Accept: application/json”

“Accept: application/ld+json”

SKOS/Schema.org Modeled Semantic Resolution (AAT example):

http://vocab.getty.edu/aat/300198841.rdf

http://vocab.getty.edu/aat/300198841.ttl

http://vocab.getty.edu/aat/300198841.nt

HTTP request of base entity endpoint (e.g., http://vocab.getty.edu/aat/300198841) with “Accept” headers for these serializations:

“Accept: application/rdf+xml”

“Accept: text/turtle”

“Accept: application/n-triples”

SPARQL Endpoints

The SKOS/Schema.org modeled Vocabularies data will still be provided at: http://vocab.getty.edu/sparql

The Linked.Art modeled Vocabularies data can be queried at:

https://data.getty.edu/vocab/sparql

Backward Compatibility

For the JSON and JSON-LD formats, users can obtain the SKOS/Schema.org modeled data by setting the HTTP header “Accept-Profile” in the following manner:

“Accept-Profile: <http://www.w3.org/2004/02/skos/core#>”

All other requests for JSON and JSON-LD will return the Linked.Art modeled version of the data.

[Getty Vocabularies LOD]

Update:

Changes to the Getty Vocabularies LOD will be happening on August 18th, 2022!

At that time, the base entities (aat, tgn, and ulan) at vocab.getty.edu will offer the Linked.Art modeled data as the default for JSON and JSON-LD serializations.

The SKOS/Schema.org modeled data will still be available at those base enitities using the "Accept-Profile" header for JSON and JSON-LD serializations.

All other serializations (RDF/XML, Turtle, N-triples) will still be modeled as SKOS/Schema.org.

The SPARQL endpoint at vocab.getty.edu will still offer the SKOS/Schema.org modeled data.

For more information, read here: http://vocab.getty.edu/ChangestoVocabLOD

[Eben English]

Hi there,

Is there still a way to retrieve the previous JSON response format? By which I mean a response that looks like:

{
  "head" : {
    "vars" : [ "Subject", "Predicate", "Object" ]
  },
  "results" : {
    "bindings" : [ {
      "Subject" : {
        "type" : "uri",
        "value" : "http://vocab.getty.edu/tgn/7017665"
      },
      "Predicate" : {
        "type" : "uri",
        "value" : "http://www.w3.org/1999/02/22-rdf-syntax-ns#type"
      },
      "Object" : {
        "type" : "uri",
        "value" : "http://vocab.getty.edu/ontology#Subject"
      }
    },
    ...
  }
}

etc etc.

Is this serialization still available anywhere? Or is it completely deprecated?

Thanks,
Eben English
Boston Public Library

[Getty Vocabularies LOD]

Yes, the previous JSON format is available for top level entities by sending the "accept-profile: <http://www.w3.org/2004/02/skos/core#>" header in your HTTP request. All other endpoints, including the the SPARQL endpoint, will give the same response as before.

What request are you trying to make?

Gregg Garcia

Software Architect

Getty Digital / J. Paul Getty Trust

[Eben English]

Hi Gregg,

I'm not getting that response from a request like:

$ curl -H "Accept: application/json" -H "Accept-Profile: <http://www.w3.org/2004/02/skos/core#>" -L https://vocab.getty.edu/tgn/1103274

or

$ curl -H "Accept: application/json" -H "Accept-Profile: <http://www.w3.org/2004/02/skos/core#>" -L https://vocab.getty.edu/tgn/7017665.json

Instead that response looks like:

{
  "http://vocab.getty.edu/tgn/1103274" : {
    "http://www.w3.org/2004/02/skos/core#altLabel" : [ {
      "value" : "Blue Grotto",
      "type" : "literal"
    }, {
      "value" : "Grotta Azzurra",
      "type" : "literal"
    } ],
    "http://purl.org/dc/terms/source" : [ {
      "value" : "http://vocab.getty.edu/tgn/source/9006541-subject-1103274",
      "type" : "uri"
    } ],

etc etc

Any advice?

Thanks,
Eben

[Getty Vocabularies LOD]

Hi, Eben.

I can see the difference now between the two formats. This is a bug we will fix.

In the meantime, you could work around the issue by performing this SPARQL query against the JSON SPARQL endpoint. I've attached a sample file with this query that can be used like so:

$ curl -X POST --data-urlencode "query@subject_query" http://vocab.getty.edu/sparql.json

Gregg

[Getty Vocabularies LOD]

Hi, Eben.

We believe this issue has been fixed. Let me know if you have any further issues.

Gregg

[Vladimir Alexiev]

Hi! Can you add these prefixes to the sample query at https://data.getty.edu/vocab/sparql-ui:

prefix crm: <http://www.cidoc-crm.org/cidoc-crm/>
prefix aat: <http://vocab.getty.edu/aat/>

In this way, the results come out a lot more compact.

Also, will there be some richer sample queries, similar to https://vocab.getty.edu/queries ?

[Getty Vocabularies LOD]

YASGUI adds the prefixes automatically when you type one into the query. Does that not work for you?

Gregg