IVOA Thesaurus

[Frederic V. Hessman]

On 31 Oct 2007, at 3:02 pm, Douglas Burke wrote:

> Frederic V. Hessman wrote:
>> Many heartfelt thanks to Doug and Alasdair!   I'll not make any  
>> changes to IVOAT until you can provide your best attempt at a  
>> rules file.
>> Does this mean that I can then routinely run cwm to produce  
>> automatically the official SKOS file with all the loose ends taken  
>> care of?
>
> I think it does, since we have a very simple set of rules to extend  
> the input RDF. However, I guess it depends on what we (or rather  
> you, as you're doing all the hard work ;-) want the format for the  
> base document of the thesaurus to be. If it's easier to work with a  
> simple text file, and then post-process that to create the RDF  
> version, then it may be better to make these additions at that  
> stage - ie go from
>
>   simple text document
>   -> simple text document + automatic clean up
>   -> RDF, N3, HTML, ...
>
> An advantage to the cwm approach is that we don't need to worry  
> about parsing, we can just concentrate on the logic. Some  
> disadvantages are that cwm can be painful to install, the syntax  
> for the rules is not standardized, and it can be awkward to do what  
> appears to be simple operations.

At this point, I hope we're not going to add much more, so a  
"settling down" phase where the loose ends get cleaned up is fine.   
If cwm will help us get a current text file version which is  
internally consistent and will warn us when additions are made if  
there are missing NT's and GT's, then we're still way ahead of where  
we are now.

Rick