Call for theory talks at Victoria interop

[Gerard Lemson]

Hi
Cc-ing the DM group as well.
Can we join the discussion on semantics and in particular the URI syntax for semantic concepts to the way vo-dml aims to use semantic vocabularies, and how they can be used in the mapping definition?
VO-DML slightly generalizes from the way vo-urp (as used in SimDM) used <<skosconcept>>, but it is otherwise pretty much the same.
In the mapping syntax we have to acknowledge that different archives may use their own terms or codes for the various concepts and we have an OPTIONMAPPING to allow annotators to indicate the proper translation. This then needs/may need a standardized way to identify the concept indeed.
Cheers
Gerard

PS
Hope this mail arrives, as my previous one did not make it to the list yet.


From: theory-bounces at ivoa.net [mailto:theory-bounces at ivoa.net] On Behalf Of Franck Le Petit
Sent: Tuesday, April 17, 2018 6:43
To: Marco Molinaro <marco.molinaro at inaf.it>
Cc: theory at ivoa.net; IVOA semantics <semantics at ivoa.net>
Subject: Re: Call for theory talks at Victoria interop

Dear Marco,
(I answer also on the Semantics mailing list)

Last week, I answered to an email by Mireille concerning the Theory vocabularies.
Theory SKOS vocabularies can be exported as .RDF files. So, it is possible to put them on the IVOA vocabulary webpage you sent us:
http://wiki.ivoa.net/twiki/bin/view/IVOA/VocabularyTestPage
Nevertheless, we will also need the present dedicated website (http://votheory.obspm.fr - excepted we would like to change its URL to make disappear the obspm.fr<http://obspm.fr> and replace it with something as “ivoa something”) since, to fully take benefit of SKOS format, we need an application above the files to search/browse in the concept list and use the relationships between concepts. This point can wait.

Apart this point, our main concern is  to get a decision concerning the syntax of URIs for concepts.
(see the discussion initiated by Nicolas Moreau on the 30th May 2017 on the Semantics mailing list).

I note that on the VocabularyTestPage, UCDs are presented in SKOS format.
So, I have a few questions since the problem we try to solve since one year, for the Theory vocabularies and concepts, is related.

1 - Namespace, dates & revisions

A concept in the SKOS UCD list seems to be only the UCD without any namespace:
Example of concept: phys.temperature.electron
(taken from: http://ivoa.net/rdf/Vocabularies/vocabularies-20091007/UCD/UCD.html)

On the other hand, vocabularies seem to be published with a date and a revision. For example:
ivoa.net/vocabularies/ucd/ucdv1.33/ucd-1.33-20180224.rdf<http://ivoa.net/vocabularies/ucd/ucdv1.33/ucd-1.33-20180224.rdf>
or  http://ivoa.net/rdf/Vocabularies/vocabularies-20091007/UCD/UCD.html

In a VO-Table in which columns are tagged with UCDs (ex: phys.temperature.electron), how do you provide the namespace that indicates to a system from which vocabulary (with revision and date) the UCD come from ?
Said otherwise, what is the rule to build the URI of a concept ? I mean the part before the # in:
http://ivoa.net/rdf/Vocabularies/vocabularies-20091007/UCD#Phys.temperature.electron

2 - Interoperability

How do you ensure the interoperability between services ?
I mean how does a system can know that:
http://ivoa.net/rdf/Vocabularies/vocabularies-2009/UCD#Phys.temperature.electron
is the same physical quantity than
http://ivoa.net/rdf/Vocabularies/vocabularies-2012/UCD#Phys.temperature.electron

Best regards
Franck



Le 17 avr. 2018 à 08:27, Marco Molinaro <marco.molinaro at inaf.it<mailto:marco.molinaro at inaf.it>> a écrit :

Hi Hervé, Carlos, all,
unfortunately there was no TCG face-to-face in Santiago, so the vocabulary topic was stuck.
I make myself a reminder and ask Mireille, so to have this discussed in the TCG in Victoria.

Mireille has prepared already material to start the discussion from, see

http://wiki<http://wiki/>.ivoa.net/twiki/bin/view/IVOA/VocabularyTestPage

Hope this helps.
Cheers,
     Marco

PS - personal opinion: it would be nice to hear your theory ideas regardless the solution to this issue.

2018-04-16 12:04 GMT+02:00 Hervé Wozniak <herve.wozniak at umontpellier.fr<mailto:herve.wozniak at umontpellier.fr>>:
Hi Carlos,

one current task for theory is to provide a clear place for its vocabulary. This is mandatory to be able to provide a revision for SimDM, to which I have committed myself.

After a long discussion on theory and semantics, launched by Marco on May 19th 2017 and suspended after Franck's mail on Oct 9th, it has been suggested (by Marco and Mireille I guess) to move the discussion at TCG level in Santiago. Indeed, any solution mixes formal (IVOA consistence) and technical (subdomain redirection) aspects. I wasn't in Santiago, so I don't know what the decisions are.

However, without any progress on this issue, it is impossible to move to a revision, even cosmetic, of SimDM.

So my participation in the Victoria meeting depends on our capacity to conclude.

Cheers
Hervé

PS: do I summarise properly the discussions in writing that 'IVOA' has to decided between:

http://(www).ivoa.net/vocabularies/theory/vocabulary#ConceptName<http://ivoa.net/vocabularies/theory/vocabulary#ConceptName>
or
http://purl.ivoa.net/theory/vocabulary#ConceptName

provided that any solution allows for redirection towards an external server for curation (e.g. purl.obspm.fr/vocab<http://purl.obspm.fr/vocab>)


Le 09/04/2018 à 19:43, Carlos Rodrigo a écrit :
Dear interop and, in particular, theory people

This is a call for contributions for the Theory session in the next IVOA Interop meeting in Victoria.

If you are going to Victoria this may and you are interested in presenting of introducing any
discussion related to theoretical data in the VO context, please, just send me your proposal.

Of course, topics related to SimDAL, SimDM, SimDAL implementation note, semantics in the theoretical
context, etc are welcome. But any new topic that you want to talk about is welcome too.

See you soon

Carlos Rodrigo
crb at cab.inta-csic.es<mailto:crb at cab.inta-csic.es>

--
Herve WOZNIAK
Astronome

Directeur-adjoint
LUPM
UMR5299 Université de Montpellier - CNRS
Place Eugène Bataillon - CC 72
F-34095 Montpellier Cédex 05

Tel/Phone : +33 4 6714 3902   Fax : +33 4 6714 4190

skype : herve.wozniak
http://herve.wozniak.fr/


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ivoa.net/pipermail/dm/attachments/20180417/854a4bde/attachment-0001.html>