GeoSciML SWG meeting Orléans 2018

• Regular business (CR, which are just typos and omission in the documentation)
  • ISO-19115-3 metadata dependency (If I recall correctly, we know have a change in schemaLocation.
  • GitHub versus SVN
  • typos, errors, omissions in document
• OWL encoding and IE
  • Right now, we have 3 axis
  • UML to RDF/OWL encoding, which means it’s an encoding exercise. Which I don’t think (happy to be convinced otherwise) is an IE.
  • "OWLification" which imply some modelling (under the assumption that the current UML does not capture all the details that could be expressed in OWL).
  • And I’m thinking here of what IGS has done with the vocabulary. It redefines the way the model is linked to the vocabulary (ie, SKOS versus full OWL model) which is a different level of commitment, keeping in mind there is a Borehole IE on the table too
  • Actual use cases of RDF à la W3C /OGC Best Practice (OneGeology use cases)
• ELFIE demonstration
• EPOS encoding

PPT: https://portal.opengeospatial.org/index.php?m=projects&a=view&project_id=420&tab=2&artifact_id=78196

Changes, errors and typos

A couple of errors in the documentation. Mineral section is missing, a property is not defined in it's own numbered section, etc..

Scott confirmed that editorial changes are simple editorial change in the document.

ParticleGeometryDescription::shape has been slated for change, but we'll wait for next major version.

Changes is schema location (?) for ISO019115-3 has been updated in the schema. They validate but not sure if the new schemas are identical to new one (the old one disappeared). Right now, we'll just wait for someone to discover and problem and address when it happens.

GitHub

GeoSciML has a GitHub account

https://github.com/opengeospatial/GeoSciML

You need a GitHub account and then link it in OGC portal and link your account to GeoSciML account. Right now Sylvain Grellet and Eric Boisvert have admin priviledge (I think).

There was discussion about moving old artefacts from previous versions under GitHub from SVN (hosted in CSIRO). It has been stressed that we should do some cleanup because SVN got cluttered over the years. Discussion need to be had with CGI/IUGS since they are "owner".

Possible issue with Enterprise Architect and GitHub. Although Github has a SVN interface, no one in the group has experience setting up a EA -> SVn -> Github environment.

Presentations

ELFIE demos (Ritchie, Grellet) TODO: add links to presentations

EPOS demo on QGIS (Grellet) TODO: add links to presentations

OWL RDF

Long discussion about JSON-LD encoding that requires RDF "schema". Some people has been working on this before

• IGS :
  • https://github.com/KoalaGeo/Geology-Ontology
• Lombardo et al,
  • https://link.springer.com/chapter/10.1007%2F978-3-319-70169-1_22
  • https://www.di.unito.it/wikigeo/index.php?title=Pagina_principale

Resolution was to aim to direct encoding of RDF as a first step to satisfy other community linked data use cases that need to refer to type definition and then re-evaluate further work.

Was agreed to engage CGI

GeoSciML paper

In Vienna, it was suggested that we write a new paper to present recent developments and update 2005 GeoSciML (Sen and Duffy). Recognise people are busy, but paper would be valuable.

Eric B. propose to write a paper layout to kick start the process and see how it goes.

-- EricBoisvert - 20 Mar 2018