Re: Axiom annotations

I've no magic up my sleeve - the aesthetic comment was intended more as 
a sigh.
There is a good case for this axiom annotations, and there is no obvious 
usable support for them in RDF backwardly compatible with OWL 1.0 
annotations. So whatever choices get made are likely to be ugly. Time to 
cover ones eyes.

Jeremy


Boris Motik wrote:
> Hello,
> 
> I completely that it isn't pretty; however, the problem is that RDF does not
> allow you to express ternary relations. In the case of the axiom
> 
> SubClassOf(Annotation(ap "bla") c1 c2)
> 
> you actually have a ternary relationship: the participants are the
> annotation, c1, and c2. Since axioms can have more than one annotation, you
> actually need a way to represent relationships of arbitrary arity. The only
> way that I know of for encoding this into RDF is to reify the relationship.
> 
> Sincerely yours,
> 
> 	Boris
> 
>> -----Original Message-----
>> From: Jeremy Carroll [mailto:jjc@hpl.hp.com]
>> Sent: 18 April 2007 10:17
>> To: Boris Motik
>> Cc: 'Alan Ruttenberg'; public-owl-dev@w3.org; 'chris mungall'
>> Subject: Re: Axiom annotations
>>
>>
>> Reification isn't pretty.
>> This seems to be only about making something work in practice.
>>
>> Jeremy
>>
>>
>>
>> Boris Motik wrote:
>>> Hello,
>>>
>>> You are right - we overlooked this point. Would you mind logging this as
>> an
>>> issue at Google Code?
>>>
>>> The problem seems to be caused by the fact that, dating from OWL 1.0,
>> you
>>> could state only binary equivalences, disjointness axioms, etc. in OWL
>> RDF.
>>> OWL 1.1 inherits this limitation from OWL 1.0 (in order to be backwards
>>> compatible); however, as you notice, this has consequences regarding
>>> annotations.
>>>
>>> There are two fixes that I see:
>>>
>>> 1. One should provide an RDF encoding that does not break up
>> equivalences;
>>> then, we might tack on the annotations appropriately.
>>>
>>> 2. A hack would be to simply repeat the axiom annotation on each
>> generated
>>> equivalence. I am not particularly fond of this, but it is a
>> possibility.
>>> Thanks for pointing this out!
>>>
>>> Sincerely yours,
>>>
>>> 	Boris
>>>
>>>> -----Original Message-----
>>>> From: public-owl-dev-request@w3.org [mailto:public-owl-dev-
>> request@w3.org]
>>>> On Behalf Of Alan Ruttenberg
>>>> Sent: 18 April 2007 01:54
>>>> To: public-owl-dev@w3.org
>>>> Cc: chris mungall
>>>> Subject: Axiom annotations
>>>>
>>>>
>>>>  From http://webont.org/owl/1.1/rdf_mapping.html
>>>>
>>>>> Axioms with annotations are reified. If s p o is the RDF
>>>>> serialization of the corresponding axiom without annotations given
>>>>> in Table 2 and the axiom contains annotations Annotation(apIDi
>>>>> cti), 1 ? i ? n, then, instead of being serialized as s p o, the
>>>>> axiom is serialized as follows:
>>>>>
>>>>> _:x rdf:type owl11:Axiom
>>>>> _:x T(apIDi) T(cti)   1 ? i ? n
>>>>> _:x rdf:subject s
>>>>> _:x rdf:predicate p
>>>>> _:x rdf:object o
>>>> What happens when an Axiom serializes as multiple triples:
>>>>
>>>>> EquivalentClasses(c1 ... cn) =>
>>>>> T(ci) owl:equivalentClass T(ci+1)   1 ? i ? n-1
>>>> -Alan
>>>
>> --
>> Hewlett-Packard Limited
>> registered Office: Cain Road, Bracknell, Berks RG12 1HN
>> Registered No: 690597 England
> 
> 
> 

-- 
Hewlett-Packard Limited
registered Office: Cain Road, Bracknell, Berks RG12 1HN
Registered No: 690597 England

Received on Wednesday, 18 April 2007 09:26:11 UTC