Re: Allowed types of punning (ISSUE-114)

You may remember that the OWL 1.1 proposal allowed unrestricted punning. 
The current inability to pun between Datatypes and Classes and the among 
Object, Data, and Annotation Properties results from perceived problems 
with the RDF serialization that allowed unrestricted punning and the 
extra mis-alignment between DL and Full that comes from this 
unrestricted punning.   A number of working group members reluctantly 
allowed the above arguments to overcome the coherent design in OWL 1.1, 
and went so far as to design the current situation.

No such technical grounds have been raised against the current version 
of punning, which allows four-way use of URIs (e.g., individual, class, 
object property, and ontology).  I do not see *any* evidence that there 
is anything left to "think through".

The current situation attempts to make the reach of OWL DL as large as 
possible with respect to OWL Full, one of the *major* driving forces in 
the development of OWL, which goes back to the initial change from 
DAML+OIL to OWL.   If this is no longer a goal for OWL, then I propose 
that we go back to the OIL situation, where the mapping into RDF was 
simply a hack to allow ontologies to be transferred over the web, and no 
alignment with the meaning of RDF graphs was envisioned.

peter


Alan Ruttenberg wrote:
> Indeed this points to further problems, as we can  pun individuals to 
> each of Object and Data properties, but not Object to Data properties.
> 
> Similarly we can pun individuals to Datatypes and to Classes, but not 
> Datatypes to Classes.
> 
> What to make of this? No three way puns, probably. Further evidence that 
> this is not thought through.
> 
> -Alan
> 
> On Jul 10, 2008, at 3:09 AM, Peter F. Patel-Schneider wrote:
> 
>>
>> Boris Motik wrote:
>>> Hello,
>>
>> [...]
>>
>>> I believe that the use cases for "individual vs. anything" type of 
>>> punning have been well documented in a number of publications.
>>> Thus, the only type of punning that could potentially be 
>>> controversial is "class or datatype vs. some type of property".
>>> Right, I don't expect people really wanting to have a property called 
>>> "xsd:integer"; however, I don't see how disallowing it makes
>>> the spec better. People can do this in OWL Full, and allowing this in 
>>> OWL DL merely allows us to handle a larger percentage of RDF
>>> graphs. Moreover, I believe that there is no distinction between the 
>>> semantics of punning in OWL Full and OWL DL; to be more
>>> precise, I don't think you can notice the difference at the level of 
>>> consequences.
>>> The same holds for punning of the form "class vs. some type of 
>>> property".
>>> Regards,
>>>     Boris
>>
>> As well, if OWL 2 allows individual/class and individual/object 
>> property punning (for example), it seems to me that it implicitly 
>> allows class/object property punning.  In fact, prohibiting the third 
>> kind of punning while allowing the first two is going to require some 
>> interesting behaviour:
>>
>> Case 1 - OK, individual/class punning:
>>
>> Declaration( NamedIndividual( ex:foo ) )
>> Declaration( Class( ex:foo ) )
>>
>> Case 2 - OK, individual/object property punning:
>>
>> Declaration( NamedIndividual( ex:foo ) )
>> Declaration( ObjectProperty( ex:foo ) )
>>
>> Case 3 - OK, both of the above categories of punning:
>>
>> Declaration( NamedIndividual( ex:foo ) )
>> Declaration( Class( ex:foo ) )
>> Declaration( ObjectProperty( ex:foo ) )
>>
>> Case 4 - not OK?????, but a subset of Case 3?????
>>
>> Declaration( Class( ex:foo ) )
>> Declaration( ObjectProperty( ex:foo ) )
>>
>> peter
>>
>

Received on Thursday, 10 July 2008 08:30:04 UTC