use of vocabulary tokens

[Rob Seaman]

Rick wrote:

> assuming that something like the revised IAU thesaurus will be  
> accepted [...], providing VOEvent with at least one set of useful  
> tokens for <Why>, it would be good for the semantic WG discussion  
> to hear how other groups would like to be able to use these tokens

The SWG does seem a bit...theoretical.

> For example, if all VOEvent wants is to standardize the <Param>'s,  
> e.g.
>
> 	<Param concept="IAU:gamma-ray_burst"/>
>
> (e.g. alongside the ucd="" usage, at least for a while), then  
> nothing needs to be done and we could have a solution very quickly.

<Params> should remain sequestered in <What>.  While this usage seems  
fine to me, it doesn't address the issue of asserting scientific  
hypotheses in <Why>.

> If, however, more complex information is needed, then something like
>
> 	<Why>
> 		<rdf:Bag>
> 			<ext-skos:NOT>
> 				<rdf:li resource="IAU:dwarf_novae"/>
> 				<rdf:li resource="IAU:gamma-ray_bursts"/>
> 			</ext-skos:NOT>
> 			<rdf:li resource="IAU:optical_outbursts"/>
> 		</rdf:Bag>
> 	</Why>
>
> becomes possible, making VOEvent explicitly semantic-web conform.   
> Or, the RDF/SKOS gobbletygook is hidden in VOEvent-specific  
> elements, implicitly, leaving
>
> 	<Why>
> 		<NOT>
> 			<Concept prob="1%">IAU:dwarf_novae</Concept>
> 			<Concept prob="2%">IAU:gamma-ray_bursts</Concept>
> 		</NOT>
> 		<Concept prob="90%">IAU:optical_outbursts</Concept>
> 	</Why>
>
> (though I'm not sure how much semantic hiding is good, if VOEvents  
> are supposed to be digestable by others).

Right.  I doubt the final syntax will closely resemble either of  
these.  I would like to see some evolution of the current <Inference>  
and <Concept> scheme, so something more like the latter.

- Rob