[OWL] [1.1] OWL 1.1 detailed proposal

[Denny Vrandecic]

No, you still use an unfitting example. SameIndividualAs does not imply 
anything in RDF-style semantics, because it is not defined in RDF. It is 
only defined in OWL.

So you still didn't bring an example where OWL needs to diverge from RDF 
semantics due to metamodelling by punning.

denny

Peter F. Patel-Schneider wrote:
> The example is an example where different results are visible to OWL depending
> on whether an RDF-style semantics is selected or a traditional first-order
> semantics is selected.  
> 
> In the former, because names denote an element of the domain of discourse, and
> the meaning of the name as a class depends on this denotion,
> SameIndividualAs(class1name class2name) implies that the two names act as
> synonyms as far as class reasoning goes.  In the latter, names separately
> denote an element of the discourse, used in individual reasoning, and a subset
> of the universe of discourse, used in class reasoning, and the two "uses" are
> completely independent.
> 
> So, perhaps it would have been better to say that 
> 	OWL 1.1. semantics will no longer match the RDF semantics.
> 
> peter
> 
> 
> From: Denny Vrandecic <dvr at aifb.uni-karlsruhe.de>
> Subject: Re: [OWL] [1.1] OWL 1.1 detailed proposal
> Date: Mon, 12 Dec 2005 19:20:25 +0100
> 
> 
>>I don't see why Meta-modelling will break RDF semantics. RDF does not 
>>have a SameIndividual-Axiom in its vocabulary, so your example doesn't work.
>>
>>Can you offer an example where RDF semantics (as opposed to RDFS 
>>semantics) gets broken with OWL Meta?
>>
>>denny
>>
>>Peter F. Patel-Schneider wrote:
>>
>>>I mean that OWL 1.1 entailment will no longer match RDF entailment (or even
>>>simple entailment), as there will be information in an OWL 1.1 document that,
>>>if translated into RDF triples, would result in divergence between OWL 1.1
>>>entailment and RDF entailment.
>>>
>>>
>>>From: Denny Vrandecic <dvr at aifb.uni-karlsruhe.de>
>>>
>>>>First, when you say Meta-modelling breaks up RDF semantics, do you 
>>>>really mean RDF? I think you rather mean RDFS?
>>
>>>>[Relevant quotes]
>>>>
>>>>
>>>>>4/ Meta-modelling and annotations
>>>>>
>>>>>Meta-ontology constructs, through punning
>>>>>	The semantics of OWL 1.1 will break the connection between classes and
>>>>>	individuals that share a name.  This diverges from the situation in
>>>>>	RDF.  The effect is that the following will be legal 
>>>>>		Class(Person)
>>>>>		Individual(Person)
>>>>>		Individual(John Person)
>>>>>		SameIndividualAs(Person Rock)
>>>>>	but that it does *not* entail
>>>>>		Individual(John Rock)

-- 
Denny Vrandecic
Institute AIFB, University of Karlsruhe (TH)
http://www.aifb.uni-karlsruhe.de/WBS/
Blog: http://semantic.nodix.net