Re: [SKOS] chatting on SKOS concepts and ontology classes (was Re: ISSUE-26: SimpleExtension proposal)

Hi Daniel,

I wonder whether we have a big problem since the beginning, because of 
the concept-centric modelling approach of SKOS, as opposed to the 
term-centric modelling approach of some thesauri (if not of most of them).

To sum up, I think SKOS concepts are partly what you call "terms" in 
thesaurus. Just go and read 
http://www.w3.org/TR/swbp-thesaurus-pubguide/#secExpressingThesaurus

The problem is that if a traditional thesaurus "Term"or "Used-for Term" 
are represented in SKOS as objects of prefLabel statements. But 
"BroaderTerm" are represented as object of skos:broader statements 
between skos:Concepts (see again 
http://www.w3.org/TR/swbp-thesaurus-pubguide/#secExpressingThesaurus).
So there is complexity here about what happened to "term" in the SKOS 
concept-centric approach. Some of the term attributes (the lexical 
values) have been treated as labels (what I call lexicalizations) while 
other have been treated as conceptual features (skos:semanticRelation 
involving skos:Concepts)

Which was not irrelevant: if you refer to a definition for term (even if 
incomplete: A "term" is a word, word pair, or word group, that is used 
in specific contexts for a specific meaning.) you see that the notion of 
term includes both a conceptual load and a lexical one. But clearly this 
choice of splitting the notion of term does not help to discuss about 
things, as we are trying to do now. Do you confirm my guess is right, or 
am I completely wrong?

By the way, please notice that when I said that "lexicalization" was 
replacing "term", it was in the scope of Guus's proposal about links 
between labels. For the general understanding of "term" I would never 
propose to replace it by "lexicalization".

I will now quote your last answers only when it is possible, trying to 
clarify the discussion
>
> And what I'm saying is that SKOS should enable people to make it 
> explicit whether they are talking about things or language--this is 
> necessary for interoperability. If you say SKOS enables this already, 
> then we should assert a best practice for people to make the 
> distinction explicit with SKOS constructs.
SKOS enables it (together with OWL which deals with the ontological 
part), yes, and guidelines shall be needed, agreed.

> I want to know  whether the "car" entity in my ontology is related to 
> the "car" lexicalization in your thesaurus.
By "entity" are you refering to "class" (the things which are the focus 
of (OWL) ontologies)?
>
>>
>> Bout don't forget that what thesauri traditionally call "terms" are 
>> in SKOS approach to be represented as instances of skos:Concept!
>
> As above, SKOS should enable people to make explicit what they are 
> talking about for interoperability--clearly, thesauri "terms" are not 
> the same as "Concepts".
Depends, cf. my remark above
>
> "broader" and "narrower" are relations that should be used to connect 
> only "terms" (aka lexicalizations). People should use different 
> relations for concepts (e.g., subclass-of, etc)
Similar remark.
And notice that subClassOf cannot be used for concepts generally. It is 
used for (OWL) classes, which are a specific aspect (the extensional 
one) of a conceptual entity. This thing will come back below.

>
>> To add to this mutual clarification process: as said, nothing like a 
>> "term" appears now in current SKOS, it only popped up in the 
>> discussion because of loosely wording used in the proposals for 
>> solving the RelationshipBetweenLabels issue. [and of course I share a 
>> great deal of responsability for that :-(] Hence my will to 
>> replace"term"by "Lexicalization" (or by anything more neutral than 
>> "term") in Guus' proposal 
>> (http://www.w3.org/2006/07/SWD/wiki/SkosDesign/RelationshipsBetweenLabels/ProposalThree) 
>>
>
> It's clear to me now that thesauri and ontologies talk about two 
> different sorts of things--lexicalizations vs things (in reality or 
> not). We need to describe these differently with SKOS.
>

>
> Yes, but as I say above, if something is a skos:Concept, it could be a 
> lexicalization or it could be an entity--two very different things.

>>
>> 1. Personnally I don't see this as a problem. It is more a feature, 
>> illustrating how SKOS and OWL can live together when it is needed.
>>
>> 2. Trying to clarify your proposal: skos:Entity would refer to the 
>> concepts that denote "real things", in the sense that they have 
>> instances, and therefore can be considered as classes in ontologies?
>
> Yes, and skos:Concept would be "lexicalizations". Perhaps we should 
> rename skos:Concept to skos:Lexicalization. Then it would be clear 
> whether people's tokens are entities or terms.
>
>> And you would have something like skos:Entity rdfs:subClassOf 
>> skos:Concept? With some rule saying that all instances of skos:Entity 
>> are also instances of owl:Class?
>
> I'd say instances of skos:Entity are instances of owl:Class. But 
> skos:Entity and skos:Concept (I prefer skos:Lexicalization) are siblings.
>
>> I don't feel it's a necessary step (because there is compatibility 
>> between OWL and SKOS, as I've demonstrated) but if you feel that this 
>> could make things a lot clearer for users of SKOS interested in 
>> ontologies (or vice versa) then I *warmly welcome it* as a proposal 
>> and am ready to help you formalizing it.
>
> I think it would make things a lot clearer. And I think widespread 
> adoption of SKOS will hinge on being clear in addition to being 
> pertinent to people's needs.
I think we rouhgly agree on the new constructs and the need to 
distinguish things, if of course we manage to get rid of this horrible 
terminology gap between us (once again, apologies for the fools who dare 
reading these mails ;-)

The only point I would object is whether skos:Entity would be disjoint 
from skos:Concept.

Just look back to this intension/extension aspect of a concept, where 
"concept" is taken in the common understanding of a "conceptual entity" 
(http://en.wikipedia.org/wiki/Extension_%28semantics%29, 
http://en.wikipedia.org/wiki/Intension)

To me, we can assume that the objects of descriptions made in SKOS (i.e. 
the skos:Concepts) are these conceptual units, and that SKOS gives means 
to describe them from the "sign" point of view (associating them to 
words, giving semanitc links and natural language definitions).
Then for some of these concepts (which implies that we have a subset of 
skos:Concept, hence a subClass of skos:Concept, and not a sibling 
class), we would have definitions relative to the extensional point of 
view, i.e. definitions refering to the objects categorized by the 
concepts. And for these special kind of concepts, SKOS may just provide 
your skos:Entity class, with nothing else: for defining the instances of 
this skos:Entity (which are owl:Classes), we can rely on OWL features, 
there is no need to be redundant.

Of course, you could prefer to say that skos:Entity refer to all the 
units that are defined somewhere as classes, and skos:Concept as the 
units that are defined by SKOS features *and not by OWL ones*, as you 
seem to advise ("skos:Entity and skos:Concept are siblings").
But making this assumption is problematic as soon as you want to 
describe with SKOS features things that are classes, which seems to me 
completely legal. You would then face the kind of inference I tried to 
illustrate before (with an instance of owl:Class being involved in 
skos:prefLabel statement, then skos:related when Tom spotted my mistake) 
that has having instances of skos:Entity inferred to be instances of 
skos:Concept, which would cause inconsistency.

Cheers,

Antoine

Received on Monday, 11 June 2007 19:32:45 UTC