Mappings (was IVOAT)
agray at dcs.gla.ac.uk
Thu Nov 29 01:58:48 PST 2007
Hi Rick, All,
It is good to see Rick include his mappings from the UCD1 vocabulary to
the IVOAT. However, I do not agree with the way in which he has linked
the two vocabularies from a ideological point of view. (I am saying
nothing about whether the terms in the two vocabularies that he has
linked makes sense or not, I am concentrating on the way in which they
Rick has declared the mappings using the skos:related property and
embedded these within the skos version of the vocabulary.
First of all, I feel that the mappings between vocabularies should be
kept completely separate from the skos version of the vocabulary. This
allows us to
1. Concentrate on generating clean versions of the vocabularies.
2. Change the mappings without needing to amend the skos version of the
3. Use different mappings in different circumstances if this meets our
Second, I feel that the skos:related property is the wrong way to relate
concepts in different vocabularies. My interpretation of the skos core,
although it is not explicitly stated, is that the properties
skos:related, skos:broader, and skos:narrower, are for defining
relationships between concepts in a single vocabulary. We should be
looking to use the SKOS mapping vocabulary specification  for this
purpose. This provides several properties for mapping between concepts
in different vocabularies. (I am in the process of declaring such
mappings between the A&A Keywords and the AOIM vocabularies. Once I have
completed this I will send it around the list.)
Alasdair J G Gray
Research Associate: Explicator Project
Computer Science, University of Glasgow
0141 330 6292
From: owner-semantics at eso.org [mailto:owner-semantics at eso.org] On Behalf
Of Frederic V. Hessman
Sent: 28 November 2007 16:55
To: IVOA semantics
The newest version of the IVOAT is now singular and it looks much nicer.
Haven't cleaned up all the minor problems found by the SKOS validator
(mostly from IAU93!) but I'm working on it as I find the time (ha!).
And, of course, I'm purposefully leaving in the Ae star/Be star/
Herbig AeBe star BT/NT cunundrum which no one has yet answered.
My version of the UCD1 docs also now includes a few linked references
to IVOAT tokens (included in the dictionary after the "prefLabel
(token)" banner), which makes for interesting comparisons of the two
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the semantics