reasoner inferences

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

reasoner inferences

Lewin Eisele
Dear list,
I'm experimenting with Protege (5.0.0 beta-17) and I'm having an issue
with inferred triples. Specifically,
in Protege I defined classes A and B and made A rdfs:subClassOf B. Then
I created an instance of class A. By
the semantics of rdfs:subClassOf I would expect a reasoner to infer that
this instance of A is also an
instance (i.e. rdf:type) of B. However, this inference does not appear
to happen and I wonder why.

I used different reasoner plugins (Hermit,Fact++,Pellet). Reasoner is
active and tick mark at "Show inferences"
is set. Everything is ticked under preferences->reasoner->displaced
inferences.

Can anyone explain?

Thanks in advance, Lewin
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: reasoner inferences

Rafael Gonçalves
Hi Lewin,

Reasoners do infer that an instance (“a”) of A is an instance of B as well. If you inspect the "Classification results” Protege view you’ll see the inferred axiom there, i.e., a Type B.  

Within the Entities/Classes/Properties tabs, Protege won’t show inferences derived via the transitivity of the SubClassOf (=>) relation, that is, if A => B, and B => C, Protege won’t show you that A => C (but it holds true nonetheless).

You can also test such things using the 'DL Query' tab, for example, to determine the types of some instance “a”, type {a} into the ‘Query (class expression)’ input box, and tick ‘Superclasses’. You should see both A and B are listed in the results, and you can press the question mark to get justification(s) for those inferences.

Hope this helps.

Cheers,
Rafael

On Aug 24, 2015, at 13:50, Lewin Eisele <[hidden email]> wrote:

Dear list,
I'm experimenting with Protege (5.0.0 beta-17) and I'm having an issue
with inferred triples. Specifically,
in Protege I defined classes A and B and made A rdfs:subClassOf B. Then
I created an instance of class A. By
the semantics of rdfs:subClassOf I would expect a reasoner to infer that
this instance of A is also an
instance (i.e. rdf:type) of B. However, this inference does not appear
to happen and I wonder why.

I used different reasoner plugins (Hermit,Fact++,Pellet). Reasoner is
active and tick mark at "Show inferences"
is set. Everything is ticked under preferences->reasoner->displaced
inferences.

Can anyone explain?

Thanks in advance, Lewin
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: reasoner inferences

Lewin Eisele

Thanks Rafael, that helped!

I also tried SPARQL Query with "SELECT ?s WHERE { ?s rdf:type :A .}" which gives me "a" as an answer. However, "SELECT ?s WHERE { ?s rdf:type :B .}" yields no answer. Can I make inferred triples also visible from within the SPARQL Query Tab?

regards, Lewin

Am 25.08.2015 um 00:45 schrieb Rafael Goncalves:
Hi Lewin,

Reasoners do infer that an instance (“a”) of A is an instance of B as well. If you inspect the "Classification results” Protege view you’ll see the inferred axiom there, i.e., a Type B.  

Within the Entities/Classes/Properties tabs, Protege won’t show inferences derived via the transitivity of the SubClassOf (=>) relation, that is, if A => B, and B => C, Protege won’t show you that A => C (but it holds true nonetheless).

You can also test such things using the 'DL Query' tab, for example, to determine the types of some instance “a”, type {a} into the ‘Query (class expression)’ input box, and tick ‘Superclasses’. You should see both A and B are listed in the results, and you can press the question mark to get justification(s) for those inferences.

Hope this helps.

Cheers,
Rafael

On Aug 24, 2015, at 13:50, Lewin Eisele <[hidden email]> wrote:

Dear list,
I'm experimenting with Protege (5.0.0 beta-17) and I'm having an issue
with inferred triples. Specifically,
in Protege I defined classes A and B and made A rdfs:subClassOf B. Then
I created an instance of class A. By
the semantics of rdfs:subClassOf I would expect a reasoner to infer that
this instance of A is also an
instance (i.e. rdf:type) of B. However, this inference does not appear
to happen and I wonder why.

I used different reasoner plugins (Hermit,Fact++,Pellet). Reasoner is
active and tick mark at "Show inferences"
is set. Everything is ticked under preferences->reasoner->displaced
inferences.

Can anyone explain?

Thanks in advance, Lewin
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user


_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: reasoner inferences

Lorenz Buehmann
Hello Lewin,

you should use the “Snap SPARQL query” plugin if you want reasoning in SPARQL.

Kind regards,
Lorenz


Thanks Rafael, that helped!

I also tried SPARQL Query with "SELECT ?s WHERE { ?s rdf:type :A .}" which gives me "a" as an answer. However, "SELECT ?s WHERE { ?s rdf:type :B .}" yields no answer. Can I make inferred triples also visible from within the SPARQL Query Tab?

regards, Lewin

Am 25.08.2015 um 00:45 schrieb Rafael Goncalves:
Hi Lewin,

Reasoners do infer that an instance (“a”) of A is an instance of B as well. If you inspect the "Classification results” Protege view you’ll see the inferred axiom there, i.e., a Type B.  

Within the Entities/Classes/Properties tabs, Protege won’t show inferences derived via the transitivity of the SubClassOf (=>) relation, that is, if A => B, and B => C, Protege won’t show you that A => C (but it holds true nonetheless).

You can also test such things using the 'DL Query' tab, for example, to determine the types of some instance “a”, type {a} into the ‘Query (class expression)’ input box, and tick ‘Superclasses’. You should see both A and B are listed in the results, and you can press the question mark to get justification(s) for those inferences.

Hope this helps.

Cheers,
Rafael

On Aug 24, 2015, at 13:50, Lewin Eisele <[hidden email]> wrote:

Dear list,
I'm experimenting with Protege (5.0.0 beta-17) and I'm having an issue
with inferred triples. Specifically,
in Protege I defined classes A and B and made A rdfs:subClassOf B. Then
I created an instance of class A. By
the semantics of rdfs:subClassOf I would expect a reasoner to infer that
this instance of A is also an
instance (i.e. rdf:type) of B. However, this inference does not appear
to happen and I wonder why.

I used different reasoner plugins (Hermit,Fact++,Pellet). Reasoner is
active and tick mark at "Show inferences"
is set. Everything is ticked under preferences->reasoner->displaced
inferences.

Can anyone explain?

Thanks in advance, Lewin
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user



_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
-- 
Lorenz Bühmann
AKSW group, University of Leipzig
Group: http://aksw.org - semantic web research center

_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user