[Pellet-users] Reasoning capabilities of Pellet

[Georgios Meditskos]

Thank you a lot  for your time (and the modified ontology).

Evren Sirin wrote:
> There are couple of issues with this ontology. The first issue is a 
> mix-up with the URI's. The default namespace and the xml:base are 
> different causing the  local names to be resolved differently and you 
> end up with two classes like this:
>
> http://www.xfront.com/owl/ontologies/camera/#ImageCapturingDeviceType
> http://127.0.0.1/ontology/extendedCamera.owl#ImageCapturingDeviceType
>
> And the ontology uses a rather unusual modeling style that causes the 
> reasoning to slow down and causes many concepts to be unsatisfiable. 
> When I ran pellet command line with -unsat option I found the 
> following concepts to be unsatisfiable:
>
> Digital, Compact, PointAndShootWithZoom, CellPhoneCamera, APS-SLR, 
> PointAndShoot, CellPhoneWithCamera, Digital-SLR, APS-Compact, SLR, 
> Digital-Video
>
> One possible problem I see is the axiom that defines NonAPS as the 
> complement of APS-Camera. Complement is a powerful construct and 
> should be used with care. Complement of APS includes things that are 
> not cameras which yield to unsatisfiability in the end. I think you 
> just want a disjointness between those classes not complement.
>
> The issue that makes reasoning slow is putting both an equivalentClass 
> and a subClassOf axiom for the same class. For example, the definition 
> of APS-Camera concept looks like this:
>
>  SubClassOf(APS-Camera Analog)
>  EquivalentClasses(APS-Camera 
> ObjectAllValuesFrom(hasImageCapturingDevice APS))
>
> Such definitions decrease the effectiveness of optimizations, 
> specificalyl absorption. Above axioms could be simplified into one 
> equivalent class axiom
>
>  EquivalentClasses(APS-Camera And(Analog 
> ObjectAllValuesFrom(hasImageCapturingDevice APS)))
>
> There are more similar examples in the ontology. I simplified couple 
> of these definitions and the classification takes just couple of 
> seconds. But there are still some unsatisfiable concepts in the 
> ontology that suggests there are more modeling issues. Maybe the 
> equivalentClass axioms for some of these concepts is too strict and 
> should be subClassOf instead. I've attached the ontology I modified if 
> you want to investigate further.
>
> Cheers,
> Evren
>
>
> On 11/5/07 4:47 PM, Georgios Meditskos wrote:
>> Hi,
>>
>> I have tried to load in Pellet 1.5.0 (using the OWL API) the 
>> ontology  below, taken from the OWL-S TC v2 collection for Semantic 
>> Web Services  discovery.
>> http://users.auth.gr/~gmeditsk/extendedCamera.owl
>>
>> The problem is that the reasoner does not terminate within a  
>> reasonable time limit.
>> In fact, I have left it for about 15 minutes and there were no  
>> results. My intention is to
>> materialize all the semantics in the beginning (using the realize()  
>> function) and then to manipulate
>> the information using Pellet's API. Is there is any combination of  
>> Pellet options
>> that might be functional for the specific ontology or the  
>> expressiveness of the ontology is to complicated
>> to be handled by Pellet?
>>
>> Thank you.
>>
>>
>> _______________________________________________
>> 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/
>>   
>