[QUAR] Re: Expressing position in RDF

[Norman Gray]

Matthew,

On 2008 Oct 16, at 21:44, Matthew Graham wrote:

>> Hi,,
>
>
>> How would you do this in FITS-WCS?
>>
>> A FITS header is a set of predicate-object pairs with, in most  
>> cases the same subject.  Put the right kind of pointy brackets  
>> around that and you've got your triples.  Your RDF problem is  
>> solved, with semantics as precisely defined (in the three Papers)  
>> as you'd want.
>
> I suggested something similar with UCDs and was told that was a bad  
> idea! Do we have a FITS ontology then that we can reference?

Yes, we have.  It's

Representations of World Coordinates in FITS, Greisen, E. W., and  
Calabretta, M. R.,Astronomy & Astrophysics, 395, 1061-1075, 2002.  
(plus Paper2 and Paper3)

It even has a URI: <http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2002A%26A...395.1061G 
 >.  Can't get more up-to-the-minute than that.

All it's missing is an <owl:Ontology> at the top and we're golden.   
How about:

@prefix : <http://fits.gsfc.nasa.gov/rdf/fits-wcs#>
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>.
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#>.
@prefix xsd: <http://www.w3.org/2001/XMLSchema#>.
@prefix owl: <http://www.w3.org/2002/07/owl#>.

<http://fits.gsfc.nasa.gov/rdf/fits-wcs> a owl:Ontology;
   rdfs:comment "The FITS-WCS ontology";
   rdfs:seeAlso <http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2002A%26A...395.1061G 
 >.

:WCS a rdfs:Class.

:crval1 a rdf:Property;
   rdfs:domain :WCS;
   rdfs:range xsd:double.
:crval2 a rdf:Property;
   rdfs:domain :WCS;
   rdfs:range xsd:double.
# ...
:pc1_2 a rdf:Property;
   rdfs:domain :WCS;
   rdfs:range xsd:double.
# ...

and so mechanically on.  You can't do much reasoning with this; but  
that's OK -- you probably wouldn't want to.

I'm perfectly serious about this, by the way.  We have a FITS  
ontology, complete in every sense apart from having the wrong type of  
puncuation.  And given that we know what to do with a one-row FITS  
table, we know structurally how to express your positions in RDF.   
Another way of putting that is to say that with Paper1 repunctuated,  
FITS becomes another serialisation (albeit not very general) of RDF.

> Otherwise all we have is a set of keyword names.

At one level, that's all that RDF is.

See you,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
Dept Physics and Astronomy, University of Leicester