RE: ISA Core Location Vocabulary

Ø  And I realize a lot of imports, and was wondering if they are all import ants...

Ha! 19111 has only a few, compared with
http://def.seegrid.csiro.au/isotc211/iso19115/2003/metadata
for example. And some other ISO ontologies are its dependants in turn -
http://def.seegrid.csiro.au/isotc211/iso19115/-2/2009/imagery
http://def.seegrid.csiro.au/isotc211/iso19119/2005/services

However, each Ontology with /code/ in the name is a 'code-list' which appears as a single class in the ISO models.
Here each is implemented as a set of individuals in a distinct graph.
We chose to do it this way because 'code-lists' are supposed to have a different lifecycle to the rest of the model.

Nevertheless, some of the other ontologies in the suite do have a lot of genuine dependencies. Look at
http://def.seegrid.csiro.au/isotc211/iso19156/2011/observation
and its dependant
http://def.seegrid.csiro.au/isotc211/iso19156/2011/sampling
for example.

This is a natural consequence of their source, which is the ISO/TC 211 'Harmonized Model', which integrates the UML from all the ISO 19100-series standards, explicitly ensuring re-use and dependency tracking.
I guess there is a question as to whether it is strictly necessary to record dependencies using owl:imports, or merely rely on use of elements within the dependencies without doing a full graph import.
I'm not sure what is considered best practice here, and would be interested to hear views.

Simon


From: Ghislain Atemezing [mailto:auguste.atemezing@eurecom.fr]
Sent: Wednesday, 8 January 2014 10:42 PM
To: Cox, Simon (CLW, Highett)
Cc: frans.knibbe@geodan.nl; public-locadd@w3.org
Subject: Re: ISA Core Location Vocabulary

Hi Simon,

[ I always get lost inside ISO documents...Thanks God we have OGC experts in this group ;) ]

The ISO ontologies you (Ghislain) found on the CSIRO site includes a skeleton with just the class hierarchy -
http://def.seegrid.csiro.au/isotc211/iso19111/2007/crs [1] but almost none of the properties, yet.
ISO/TC 211 will complete these in due course.

Yes, I can see it now. And I realize a lot of imports, and was wondering if they are all import ants...(see http://www.essepuntato.it/lode/owlapi/http://def.seegrid.csiro.au/isotc211/iso19111/2007/crs<http://www.essepuntato.it/lode/owlapi/http:/def.seegrid.csiro.au/isotc211/iso19111/2007/crs> )

Please note that the ISO ontologies are/will be very literal translations of the UML models, so may miss a few tricks that would be used starting with OWL, but do nevertheless embody a great deal of analysis and 'tradition'.


Regarding UMLs, it seems to me that the import rules strictly follow the UML model packages and dependencies (Figure 4, Page 15) of the document at [1] . Am I wrong?  With the folks at IGN-France, the first version of the CRS vocabulary we worked on take the core elements within the UML diagram in Page 22,
Figure 6 named "SC_CoordinateReferenceSystem package. (see the attached file)[cid:image002.jpg@01CF0D1F.90038210]
Some harmonization would probably be desirable.

+1 . And this can be done within this CG?

Best,
Ghislain

[1] portal.opengeospatial.org/files/?artifact_id=39049