DCAT profle of ISO 19115-1

[MortenBorrebaek]

In the AHG on e-Government recommendations were made for the future revision of ISO standards and new standard projects:
-ISO/TC 211 should continue to derive RDF/OWL from the harmonised UML models in the ISO/TC 211 standards. Existing activities on revising mapping rules from UML to OWL/RDF will improve the ontology and mapping rules should be an ongoing
activity, according to experience from the communities. In addition, ISO/TC 211 should link to existing vocabularies and extend other vocabularies applied in e-Government with our spatial extensions where needed.

-In the revision cycles of ISO/TC 211 standards, focus should be given to investigate if there are other and more generic IT standards which it would be relevant to build geospatial extensions on top of. Of particular focus are standards from ISO/IEC
JTC1 and W3C, which are most focused in e-Government communities. This comes as an extension to the cooperative work between ISO/TC 211 and OGC .

For ISO 19115-1 the following was discussed:

Within eGovernment DCAT is one of the most implemented standards (at least in Europe). We see that many countries (at least in Europe) is setting up DCAT implementations, and the European Data Portal (EDP) applies DCAT.

A mapping from ISO 19115 metadata mandatory in the European INSOIRE directive to DCAT is provides, called GeoDCAT.

The challenge is that ISO 19115-1 fundamentals contains more metadata than what is specified in DCAT/GeoDCAT.
In the digital transformation metadata is considered as most important, to evaluate if data is applicable for decision making, and ISO 19115-1 definitely contribute to such an approach.

It should be considered to provide a DCAT implementation of metadata specified in ISO 19115-1 .

[PeterParslow]

(commenting here as lead editor of UK's "GEMINI" - profile of ISO 19115:2003)

Here's the UK's proposed mapping from our national subset of ISO 19115 to DCAT: agiorguk/gemini#41. It is largely based on work done by W3C: https://www.w3.org/2015/spatial/wiki/ISO_19115_-_DCAT_-_Schema.org_mapping

See also Andrea Perego's mapping at https://github.com/GeoCat/iso-19139-to-dcat-ap/blob/master/documentation/Mappings.md

[PeterParslow]

One of the issues with such a mapping is that DCAT & 19115 take a different approach to "distribution". In DCAT, a lot more things are "per distribution" rather than "per dataset / metadata record".

One particular issue (noted in the current OGC Code Sprint on OGC API Records) is that DCAT describes the license & access rights as a property of the distribution. In ISO 19115 the constraints & restrictions are in DataIdentification & there is no way to tie multiple constraint sets to multiple distributions.

[PeterParslow]

Thoughts from November 2024 plenary week WG7 discussion. Should this be:

• a TR mapping ISO 19115-1:2014 to / from DCAT?
• a TR after the revision of ISO 19115-1 which is about to start?
• something else, e.g. a IWA?

[PeterParslow]

This looks to be a pretty thorough mapping: https://docs.google.com/spreadsheets/d/1pJkKgGa655Dv06_UFwzeYoje9hEwzifE/edit?usp=sharing&ouid=115738132131001964515&rtpof=true&sd=true

Developed in the GeoNetwork team & shared with me by Jeroen Tichler during the OGC Metadata code sprint, with implicit encouragement to reuse (need to check whether that was "just" in the context of the code sprint), Nov 2024, suggesting that François Prunayre would be better positioned to explain the intricacies of the spreadsheet.