RE: SKOS & SIMILE, concepts, terms, URIs, mappings

Hi Mark, good to hear from you.

> 
> > The reason why I previously didn't think expressing lexical 
> mappings in
> RDF
> > was really necessary was because they could be 
> automatically discovered by
> a
> > computer process at any time.  But then I guess storing and 
> publishing the
> > results of that process could be useful, especially when 
> operating on
> large
> > data sets. 
> 
> I think we do need to do this, in fact I think this is related to our
> previous discussion on giving URIs to altLabels?

Yes I was thinking of this when I made these comments.  (I have some ideas
about how to express lexical mappings that does not require the altLabels to
have their own URI.)  

> Automated discovery doesn't quite work in the way you 
> describe, because the
> computer will generate lexical mappings, but then anywhere 
> between 10% and
> 50% of them are incorrect. So you need some way of capturing them, and
> reviewing them by hand if necessary.
> 
> Does that seem sensible?

Yes absolutely.  Could you possibly provide me with a list of all the types
of useful lexical mapping you can think of?  

This is splitting hairs a little bit, but I think it would be more accurate
to say that a lexical mapping may or may not reflect a close semantic
mapping.  So a lexical mapping is never 'incorrect' if it captures some sort
of lexical similarity between labels, even if there is no semantic mapping
between the corresponding concepts. 

(This is just my attempt to make a very clear distinction between the world
of lexical comparison of character strings, and the world of semantic
comparison of intended meaning.)

Yours,

Alistair.

Received on Tuesday, 8 June 2004 08:47:33 UTC