W3C home > Mailing lists > Public > public-esw-thes@w3.org > May 2006

Re: altLabels in different langauges

From: Mark van Assem <mark@few.vu.nl>
Date: Wed, 31 May 2006 18:18:31 +0300
Message-ID: <447DB3C7.5030101@few.vu.nl>
To: Antoine Isaac <Antoine.Isaac@KB.nl>
CC: Alistair Miles <a.j.miles@rl.ac.uk>, public-esw-thes@w3.org

Hi Antoine, Alistair,

> Isn't that transferring to the annotations the complexity the
> "Terms-as-classes" solution was blamed for?

We have had this exact discussion before... see [1] and [2].

This solution repeats the label, introducing redundancy. It is accepted 
practice to avoid redundancy because redundant elements makes it easier 
for update errors to occur (label changed in one place, forgotten to be 
changed in the other). And quoting myself [2]: 'I think this is a 
solution, but in principle this method could be used everywhere you 
normally use a class to group information about an entity. I think the 
more usual way to do this in RDF or OWL is introduce a class.'

> I see 2 possible problems:
> - one obvious is the complexity of the thing for the humans involved in
> the loop. Of course we always claim that this is not the core of
> semantic web concerns, but a model which is cognitively inadequate will
> be difficult to adopt.

I agree this is a concern in general.

But in terms of complexity SKOS is much closer to FOAF (arguably the 
most used schema and probably one of the simplest schemas we can expect 
to appear on the SW) than to e.g. (Onto)WordNet, or e.g. the toy-example 
Wine ontology [3]. If we cannot explain people how to correctly use SKOS 
- only 5 classes and 26 properties - then we can just quit with the 
Semantic Web now anyway.

And I think that we can expect those who will do the conversions to be 
knowledgeable and/or motivated to learn SW stuff. If they do not see the 
value of a SKOS version of their thesaurus, why would they be unwilling 
to take the time to thoroughly understand SKOS anyway?

> in reasoning engines. What if you want state that your translation is
> symmetric (if X has Y as a translation, Y has X as a translation)? You

Why couldn't you use owl:SymmetricProperty ?

> when the model is created. Similar problems could happen if you want to
> create transitive 'links' between your terms (perhaps not valid for

Why couldn't you use owl:TransitiveProperty ?

What is wrong with adding the triples yourself if you don't use OWL? You 
have to do that in RDF(S)anyway. Although I think it is usual in the RDF 
community that the fact which props are transitive is meta-knowledge, 
and the applications simply have built-in transitivity rules for those 
props. E.g. hypernymy in WordNet.

In any case, what you mention could be limiting factors on the 
usefulness of the Term approach, but not hinder its basic usefulness as 
outlined in [2].


  Mark F.J. van Assem - Vrije Universiteit Amsterdam
        markREMOVE@cs.vu.nl - http://www.cs.vu.nl/~mark
Received on Wednesday, 31 May 2006 15:19:38 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 2 March 2016 13:32:07 UTC