Vocab AND Ontology?

[Tony Linde]

Thanks for the response, Andrea. I'll clarify what I was aiming for. This
was to work only on an SV: the first draft so that the wg can confirm that
the 'OWL with derived SKOS etc' approach will work; subsequent drafts up to
the next interop to allow people to build applications for the SV and
experiment with extending the OWL-based SV into an ontology. It is not
mandatory that WDs proceed to REC. If, at the next interop, the wg decides
that only the ontology will be developed and any SV will always be a derived
entity (so the REC document only relates to the ontology and how to derive
vocabs), then the SV drafts will be abandoned/changed and an ontology draft
begun. Or it could be that both formats are wanted to be developed in
parallel, as you suggest. Or some other solution entirely: I just wanted to
get some trial developments out so people could work with them.

> final goals in terms of IVOA Recs.  If we agree that what we need is

I hope the above clarifies that point: the only immediate goal is the next
interop and SV drafts in the leadup to that. We'll set 'final' goals then.

> Second. It is clear from the discussion that the approach proposed in
> the draft is not convincing the majority of the active members of the
> WG. I'm not referring here to the format (XML or other) but to the
> content of the SV: tokens to be assembled in an UCD-like way to
> express astronomical concepts.

I think that is right. That is why I suggested the parallel discussions on
those points but based on the understanding that concepts will be OWL items
from which SV terms can be derived using a small number of relationships.
But there is still more to discuss, I think.

> we whould probably want to formalize this first step within the IVOA.

As I said above, I think it is a little too soon to formalise anything. The
IVOA as a whole can know that we are working on an OWL-based SV from which a
SKOS-based vocabulary will be derived and that drafts will be made available
as soon as possible for people to work on but that a final decision on
formats will be made at the next interop. I'm pretty sure this will be
agreeable to the TCG.

T.

> -----Original Message-----
> From: owner-semantics at eso.org [mailto:owner-semantics at eso.org] On
> Behalf Of Andrea Preite Martinez
> Sent: 24 September 2007 11:07
> To: semantics at ivoa.net
> Subject: Re: Vocab AND Ontology?
> 
> 
> > I'd like to propose, if it is feasible, a way forward on the issue of
> > creating an SV.
> 
> I think that Tony's effort to reconcile the different points of view
> emerging from the discussion, and at the same time to propose a
> practical way to go on working, is greatly appreciable.
> I see only a couple of points to clarify to remove possible
> ambiguities.
> 
> First, I think we should shape and trim our effort having in mind not
> only the what, who and how addressed by Tony, but also which are our
> final goals in terms of IVOA Recs.  If we agree that what we need is
> both a Standard Vocabulary and an Ontology, we probably have to think
> of two subsequent but distinguishable IVOA documents. This just out of
> methodology. This even in the case where the list of concepts
> (Ontology) should coincide with the list of SV terms. This even in the
> case we thought of the SV as a sort of by-product of the Ontology.
> 
> Second. It is clear from the discussion that the approach proposed in
> the draft is not convincing the majority of the active members of the
> WG. I'm not referring here to the format (XML or other) but to the
> content of the SV: tokens to be assembled in an UCD-like way to
> express astronomical concepts. The advantages were (are! :-) ) :
> a) we know already how to deal with UCDs; b) slower rate maintainance
> / up-grade; c) the proposal is there, the SV could be used now.
> The main (if not the only) disadvantage was (is) the absence of
> ontological structure (obvious, they are tokens, not concepts!) and
> the need to start all over again, should we decide to go for a
> Vocabulary and then an Ontology.
> 
> Changing tokens into concepts, using as our starting point the IAU
> Thesaurus has no great disadvantages, because the Thesaurus can be
> rapidly up-graded with a handful of concepts, and we have already the
> basic relationships BT, NT, etc. if we want to go on with an Ontology.
> 
> I think to correctly interpret the feelings of the authors of the
> draft (please correct me if I got them wrong!) saying that we don't
> object changing terms into concepts. It can even be simpler for
> users/tools of other WGs to use them.
> 
> As Tony said, a group appointed by the WG can then write a new draft,
> say on "The IVOA Standard Vocabulary" in one or two months.
> We will be only on the first step of the ladder, to cite Norman, and
> we whould probably want to formalize this first step within the IVOA.
> If we think we are satisfied with this, we can stop here and talking
> about formats could be a waste of time.
> 
> If, on the other hand, we think we have to climb the ladder (I
> personally think so) we can at the same time put together people and
> efforts around another draft, and start climbing further. With no
> hurry. There wan't be the risk that other WGs define their own SV, and
> the IVOA being tranformed into a Tower of Babel. The SV will be
> already there, in its status of WG-agreed, Rec-to-come document.
> 
> Cheers,
> Andrea
> 
> 
> =======================================================================
> ============
> Andrea Preite Martinez                 andrea.preitemartinez at iasf-
> roma.inaf.it
> IASF                                   Tel.IASF:+39.06.4993.4641
> Via del Fosso del Cavaliere 100        Tel.CDS :+33.3.90242452
> I-00133 Roma                           Cell.   :+39.320.43.15.383
>                                         Skype   :andrea.preite.martinez
> =======================================================================
> ============
>