[Pellet-users] One bug and one strange observation -- please help

[Evren Sirin]

Hi Sebastian,

The first issue you describe regarding missing superclass is due to a 
bug in the internals of Pellet. I've recorded this issue in the tracker [1].

You are right that getEquivalentClasses(B) function call in Pellet 
deliberately does not return B itself. This is how the internal Pellet 
API behaves but we change the behavior of such functions depending on 
the interface we implement. For example, similar query through Jena 
interface will return B itself because that is how the behavior is 
described in the Jena javadocs (similarly the query for subclasses also 
includes the class itself). The OWLAPI javadocs for these functions are 
not detailed enough so Pellet and FaCT++ implements it differently. I'll 
clarify with Matthew regarding what is the expected behavior in OWLAPI.

Cheers,
Evren

[1] http://cvsdude.com/trac/clark-parsia/pellet-devel/ticket/115


On 4/7/08 11:51 AM, Sebastian Brandt wrote:
> Hello there,
>
> I am using Pellet to evaluate DL queries in a JAVA program. In this 
> context, I have observed that Pellet answers some queries incorrectly. 
> I would be glad to get some feedback, maybe I am using Pellet in the 
> wrong way. I have prepared a tiny ontology and some screenshots of the 
> DL query pane in Protege 4 to demonstrate the problems. There are two 
> queries for which, it seems to me, Pellet does not evaluate the query 
> correctly:
>
> In my ontology, there's a class B, subclass of Thing, that doesn't 
> have any additional information asserted on it (or inferred either). 
> The first query asks for super classes of the class expression
>
>   (B and hasInt some int),
>
> where hasInt is a datatype property and int the OWL datatype. FaCT++ 
> correctly returns B as a super class but Pellet doesn't. I am sure 
> this is not as it should be. The following pictures show first the 
> output of FaCT++ (Picture 8) and then the output of Pellet (Picture 9) 
> in the Protege DL query pane.
>
>
> ------------------------------------------------------------------------
>
>  
>  
>
> ------------------------------------------------------------------------
>
>
>
>
> The second query asks for equivalent classes of B and super classes of 
> B. The first picture (Picture 6) shows the answer given by FaCT++: B 
> is equivalent to B and Thing is a super class. Pellet (Picture-7) does 
> not find B as an equivalent class of B. Is this intentionally left out 
> (i.e., interpreting the query as: any class equivalent to BUT 
> DIFFERENT FROM class B)?
>
>
> ------------------------------------------------------------------------
>
>  
>  
>
> ------------------------------------------------------------------------
>
>
>
> I have attached the ontology to this email although the example is so 
> simple it might be easier for you to add a new subclass of Thing and a 
> datatype property to any other ontology. Any hints as to why this is 
> as it is are greatly appreciated. Thanks a lot for your help. Sebastian.
>
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Pellet-users mailing list
> Pellet-users at lists.owldl.com
> http://lists.owldl.com/mailman/listinfo/pellet-users
> _______________________________________________
>
> Sponsored by Clark & Parsia, LLC http://clarkparsia.com/