SKOS concepts in VOTable

[Norman Gray]

Dave, hello.

Thanks for your comments.

On 2012 Jun 1, at 15:52, Dave Morris wrote:

> The current suggestions
> 
>    <someVOElement ...>
>        <link content-role='type' href='http://www.ivoa.net/rdf/IAUT93#Quasar' />
>        <link content-role='type' href='http://www.ivoa.net/rdf/UCD#src.redshift' />
>        <link content-role='type' href='http://www.ivoa.net/myStandard/my-favourite-utype' />
>    </someVOElement>
[...]
> Both look fairly similar to the solution we adopted in VOSpace-1.0.
> 
>    <node ...>
>        <properties>
>            <property type='ivo://ivoa.net/semantics/core#iau.rdf'>http://www.ivoa.net/rdf/IAUT93#Quasar</property>
>            <property type='ivo://ivoa.net/semantics/core#ucd.rdf'>http://www.ivoa.net/rdf/UCD#src.redshift</property>
>            <property type='ivo://ivoa.net/semantics/core#utype'>http://www.ivoa.net/myStandard/my-favourite-utype</property>
>        </properties>
>    </node>

If I'm reading you correctly, I think the effective difference is that the @type attribute in property indicates a fine-grained relationship between the <node> and the property value, whereas the only relationship in the VOTable LINK suggestion is the rather coarse relationship of 'this VOTable element has type...'.  In both cases, there's no restriction on the vocabulary used -- LINK/@href can be anyURI.

> For details of how the VOSpace properties are defined see
> 
>    http://www.ivoa.net/Documents/VOSpace/20111202/PR-VOSpace-2.0-20111202.html#sec3_2

Aha.  I think I see the (slight) conceptual difference. The VOSpace property mechanism is for associating generic properties with nodes, so the role of the property/@type attribute is to say what the type of the relation is, quite generally (that is, it's a predicate, in RDF terms).

The VOTable case is a more restricted one.  Here, the only relationship we're discussing is the 'has type' relation which @content-role='type' creates, between the element containing the <LINK>, and the value of LINK/@href.

So yes, as you suggest, these two would become equivalent if LINK/@content-role were changed from NMTOKEN to anyURI.  I broadly agree with you.  The only downside I see with that is that it requires a schema change, which I think the VOTable WG are reluctant to countenance (though does anyone actually _use_ LINK/@content-role for anything? Would anyone even _notice_ a schema change here? (I suspect the answer to the second question is yes...)).

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK