Data and metadata model
Caliper underlying data model is RDF, a W3C Recommendation, used in conjunction with a handful of public RDF vocabularies such as SKOS (used to express hierarchical structures), XKOS (to express statistic-specific constructs in SKOS resources), and metadata schemes such as FOAF, VoiD, DCAT, Dublin Core. When needed, OWL ontologies are specified.
Caliper relies on open sources technologies.
Browsing: We use three tools for displaying, browsing and searching against the repository of classifications and correspondences:
- SKOSMOS : https://github.com/NatLibFi/Skosmos/wiki. Skosmos is fed by Fuseki.
- PMKI is the read-only version of VocBench, it shows exactly the same data shown in VocBench, stored in GraphDB.
- Drupal ingests the RDF accessing the Sparql endpoint of Fuseki, it stores it internally and it visualizes it according to our internally defined views.
Each of these tools also support search.
- GraphDB, is the triple store behind VocBench.
- Fuseki, the triple store used for publication of data. It feeds Skosmos, Drupal, Loddy, and it provides the SPARQL Endpoint also used by Drupal to populate the Caliper website.
Serving LOD: Loddy, for content negotiation: https://bitbucket.org/art-uniroma2/loddy/downloads/Loddy_DEMO_2016-03-01.zip
Download: data for download are generated by means of internal routines, then made available to the public in Drupal.
- VocBench, the editing software we use:
Online query: YASGUI provides a convenient user interface to the SPARQL endpoint provided by Fuseki.
Caliper website is build in Drupal.
Tools in a lifecycle
The picture below shows the main tools used in Caliper, connected by arrows to sketch the data flow across the tools. Tools names are in black text, the functionality they provide in blue.