Golden Agents Datasprint 12 March

Introduction

Written here is some insight information in the way the data is stored and can be queried.

Overview of datasets used in this datasprint

Internal (hosted in the data.create.humanities.uva.nl triplestore)

Adamlink (graph: https://data.create.humanities.uva.nl/id/adamlink/)
- Buildings (graph: https://data.create.humanities.uva.nl/id/adamlink/buildings/)
- Districts (graph: https://data.create.humanities.uva.nl/id/adamlink/districts/)
- Persons (graph: https://data.create.humanities.uva.nl/id/adamlink/persons/)
- Streets (graph: https://data.create.humanities.uva.nl/id/adamlink/streets/)
ECARTICO (graph: https://data.create.humanities.uva.nl/id/schrijverskabinet/)
ONSTAGE (graph: https://data.create.humanities.uva.nl/id/schrijverskabinet/)
Rijksmuseum (temporary) (graph: https://data.create.humanities.uva.nl/id/rijksmuseum/)
Schrijverskabinet (graph: https://data.create.humanities.uva.nl/id/schrijverskabinet/)

Externally accessible data in triplestores

KB (STCN, DBNL, and more) (hosted and queryable at http://data.bibliotheken.nl/sparql)
RKD Artists (hosted and queryable at https://data.netwerkdigitaalerfgoed.nl/rkd/rkdartists/)
Wikidata (see https://query.wikidata.org/sparql)

Futher reading?

If you're not familar to the terms and building blocks of the semantic web, then reading some of these guides and websites might help to quickly get you up to speed:

The VU's Digital Humanities Workbench has more information, a video, and other references on this topic.
Take a look at the Wikidata SPARQL Tutorial.
The website Programming Historian has an introductory walkthrough to get you started with the principles of Linked Open Data.

The CREATE triplestore

An ordinary SPARQL endpoint can be found on /sparql, but you may want to use the yasgui SPARQL editor on the landing page.

Datasets created and/or published by CREATE are (at the time of writing) stored, sparqlable and browsable on https://data.create.humanities.uva.nl. The interface of this tool serves the properties and relations of each entity (each expressed as URI) in a human readable format. The example in the image below gives all information on Arnoud van Halen, who can be identified by URI https://data.create.humanities.uva.nl/id/schrijverskabinet/person/arnoud-van-halen (link) in the dataset of the Panpoëticon Batavûm dataset.

We state that Arnoud van Halen can be identifier by this URI, which is also a URL, and that he is of a rdf:type (abbreviated a) schema:Person (cf. the schema vocabulary). More information on e.g. his birthdate or birthplace are stored in the respective properties and values schema:birthDate and schema:birthPlace.

This is a visual way to display the actual data that is stored as triples in the datastore. The same information on this person can be read in a very readable RDF serialization [=saved textual form] that is called turtle or ttl:

@prefix schema: <http://schema.org/> .

<https://data.create.humanities.uva.nl/id/schrijverskabinet/person/arnoud-van-halen> a schema:Person ;
    schema:birthDate "1673"^^xsd:gYear ;
    schema:birthPlace _:amsterdam ;
    schema:deathDate "1732"^^xsd:gYear ;
    schema:deathPlace _:amsterdam ;
    schema:disambiguatingDescription "Schilder, graveur, dichter en verzamelaar" ;
    schema:mainEntityOfPage <http://www.schrijverskabinet.nl/portret/arnoud-van-halen/> ;
    schema:name "Arnoud van Halen" ;
    schema:subjectOf <http://www.schrijverskabinet.nl/artikel/arnoud-van-halen/>,
        <https://data.create.humanities.uva.nl/id/schrijverskabinet/artwork/32> ;
    owl:sameAs <http://data.bibliotheken.nl/id/dbnla/hale001> ;
    foaf:depiction "http://www.schrijverskabinet.nl/wp-content/uploads/2016/05/Van-Halen-SK-A-1738-2-afwijkend-formaat-2.jpg" .

All urls between <> can be considered URIs and refer to other entities. This can be an external website, such as the value of the schema:mainEntityOfPage property, or another URI within the scope of this dataset: https://data.create.humanities.uva.nl/id/schrijverskabinet/artwork/32. Each of these entities have their own block in the data:

@prefix schema: <http://schema.org/> .

<https://data.create.humanities.uva.nl/id/schrijverskabinet/artwork/32> a schema:VisualArtwork ;
    schema:about <https://data.create.humanities.uva.nl/id/schrijverskabinet/person/arnoud-van-halen> ;
    schema:artist <https://data.create.humanities.uva.nl/id/schrijverskabinet/person/arnoud-van-halen> ;
    schema:name "Portret van Arnoud van Halen"@nl .

In this case, this entity is of a rdf:type schema:VisualArtwork (cf. the schema vocabulary). The painting is schema:about Arnoud van Halen, but is also made by him (cf. schema:artist). It thereby follows from the data that this is a self-portrait.

All data in RDF follow this same structure of <object> <predicate> <subject> and each of these are written as URI, either as URL (e.g. https://example.org), or URN (urn:identifier:12345).

Sparql

SPARQL queries can we written directly on the data.create.humanities.uva.nl page in the yasgui (Github) sparql editor. This tool helps to structure, check, and save your SPARQL queries, and presents a nicely formatted table as result.

A list of usefull example queries can be found at /queries

Getting started