Quantcast

A query regarding inconsistency in Protege Ontology!

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

A query regarding inconsistency in Protege Ontology!

Aparna Lalingkar
Hello Everybody,

I have created an ontology for describing applet-based mathematics tasks.

While creating the ontology I have followed all the syntax rules and avoided the same named classes etc. I am using protege 3.5 just for creation of ontology and then after freezing it will import it in Protege 5.0

So far I have 185 Classes (Class and sub-class together), Total Properties 99 (69 object properties out of which 42 are properties with inverse specified, 31 data type and 8 annotation)

I decided to check the consistency just to be on safer side.So I used Pallet as a reasoner (which is inbuilt we only need to select it). The consistency check report shows all classes inconsistent.

Can anybody tell me why is it so? Do I need to import it in Protege 5.0 and then use Pallet reasoner to check the consistency?

Thanks and regards,

Dr Aparna Lalingkar
MERI, University of Haifa, Israel

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

Re: A query regarding inconsistency in Protege Ontology!

Michael DeBellis-2
The thing about logic and theorem provers (which is essentially what reasoners are) is that one error can make the whole ontology inconsistent.  If you remember from logic you can derive anything once you have an inconsistent axiom like p & not(p). So just because the whole ontology looks red it may be just one error.

Yes you absolutely should do everything in protege 5, not build in an older version and test in the latest. Also, IMO a best practice is to run the reasoner a lot. I run it after almost every change.  Btw, IMO that's a best practice for any software development, it's part of Agile methods for example, do testing more or less in parallel with development, not all at once after you finish development as in Waterfall models. That way you find problems earlier when they are easier and less expensive to fix.

As for debugging your ontology you should see a "?" icon next to the message that says you have an error. If you click on that icon the reasoner should give you the explanation. Or if you just carefully read the output of the reasoner sometimes that will explain it. If you still can't figure it out it's best to include your ontology and the error message so we can help explain it. Hope that helps.

Michael

> On Apr 15, 2017, at 12:45 PM, Aparna Lalingkar <[hidden email]> wrote:
>
> Hello Everybody,
>
> I have created an ontology for describing applet-based mathematics tasks.
>
> While creating the ontology I have followed all the syntax rules and avoided the same named classes etc. I am using protege 3.5 just for creation of ontology and then after freezing it will import it in Protege 5.0
>
> So far I have 185 Classes (Class and sub-class together), Total Properties 99 (69 object properties out of which 42 are properties with inverse specified, 31 data type and 8 annotation)
>
> I decided to check the consistency just to be on safer side.So I used Pallet as a reasoner (which is inbuilt we only need to select it). The consistency check report shows all classes inconsistent.
>
> Can anybody tell me why is it so? Do I need to import it in Protege 5.0 and then use Pallet reasoner to check the consistency?
>
> Thanks and regards,
>
> Dr Aparna Lalingkar
> MERI, University of Haifa, Israel
> _______________________________________________
> 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
|  
Report Content as Inappropriate

Re: A query regarding inconsistency in Protege Ontology!

Aparna Lalingkar
Many thanks Michael! I will again try to debug.

Regards,

Aparna

On Mon, Apr 17, 2017 at 5:06 PM, Michael DeBellis <[hidden email]> wrote:
The thing about logic and theorem provers (which is essentially what reasoners are) is that one error can make the whole ontology inconsistent.  If you remember from logic you can derive anything once you have an inconsistent axiom like p & not(p). So just because the whole ontology looks red it may be just one error.

Yes you absolutely should do everything in protege 5, not build in an older version and test in the latest. Also, IMO a best practice is to run the reasoner a lot. I run it after almost every change.  Btw, IMO that's a best practice for any software development, it's part of Agile methods for example, do testing more or less in parallel with development, not all at once after you finish development as in Waterfall models. That way you find problems earlier when they are easier and less expensive to fix.

As for debugging your ontology you should see a "?" icon next to the message that says you have an error. If you click on that icon the reasoner should give you the explanation. Or if you just carefully read the output of the reasoner sometimes that will explain it. If you still can't figure it out it's best to include your ontology and the error message so we can help explain it. Hope that helps.

Michael

> On Apr 15, 2017, at 12:45 PM, Aparna Lalingkar <[hidden email]> wrote:
>
> Hello Everybody,
>
> I have created an ontology for describing applet-based mathematics tasks.
>
> While creating the ontology I have followed all the syntax rules and avoided the same named classes etc. I am using protege 3.5 just for creation of ontology and then after freezing it will import it in Protege 5.0
>
> So far I have 185 Classes (Class and sub-class together), Total Properties 99 (69 object properties out of which 42 are properties with inverse specified, 31 data type and 8 annotation)
>
> I decided to check the consistency just to be on safer side.So I used Pallet as a reasoner (which is inbuilt we only need to select it). The consistency check report shows all classes inconsistent.
>
> Can anybody tell me why is it so? Do I need to import it in Protege 5.0 and then use Pallet reasoner to check the consistency?
>
> Thanks and regards,
>
> Dr Aparna Lalingkar
> MERI, University of Haifa, Israel
> _______________________________________________
> 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
|  
Report Content as Inappropriate

Re: A query regarding inconsistency in Protege Ontology!

Aparna Lalingkar
Hello Michael,

Your tip was helpful. I imported it to the protege 5.0 and ran the reasoner. It showed the one inconsistency that was making whole ontology Red. I could correct it. Again ran the reasoner and this time it did not give any error message and it continued running. So I hope this means my ontology is now consistent.

Regards,

Aparna

On Mon, Apr 17, 2017 at 10:16 PM, Aparna Lalingkar <[hidden email]> wrote:
Many thanks Michael! I will again try to debug.

Regards,

Aparna

On Mon, Apr 17, 2017 at 5:06 PM, Michael DeBellis <[hidden email]> wrote:
The thing about logic and theorem provers (which is essentially what reasoners are) is that one error can make the whole ontology inconsistent.  If you remember from logic you can derive anything once you have an inconsistent axiom like p & not(p). So just because the whole ontology looks red it may be just one error.

Yes you absolutely should do everything in protege 5, not build in an older version and test in the latest. Also, IMO a best practice is to run the reasoner a lot. I run it after almost every change.  Btw, IMO that's a best practice for any software development, it's part of Agile methods for example, do testing more or less in parallel with development, not all at once after you finish development as in Waterfall models. That way you find problems earlier when they are easier and less expensive to fix.

As for debugging your ontology you should see a "?" icon next to the message that says you have an error. If you click on that icon the reasoner should give you the explanation. Or if you just carefully read the output of the reasoner sometimes that will explain it. If you still can't figure it out it's best to include your ontology and the error message so we can help explain it. Hope that helps.

Michael

> On Apr 15, 2017, at 12:45 PM, Aparna Lalingkar <[hidden email]> wrote:
>
> Hello Everybody,
>
> I have created an ontology for describing applet-based mathematics tasks.
>
> While creating the ontology I have followed all the syntax rules and avoided the same named classes etc. I am using protege 3.5 just for creation of ontology and then after freezing it will import it in Protege 5.0
>
> So far I have 185 Classes (Class and sub-class together), Total Properties 99 (69 object properties out of which 42 are properties with inverse specified, 31 data type and 8 annotation)
>
> I decided to check the consistency just to be on safer side.So I used Pallet as a reasoner (which is inbuilt we only need to select it). The consistency check report shows all classes inconsistent.
>
> Can anybody tell me why is it so? Do I need to import it in Protege 5.0 and then use Pallet reasoner to check the consistency?
>
> Thanks and regards,
>
> Dr Aparna Lalingkar
> MERI, University of Haifa, Israel
> _______________________________________________
> 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
|  
Report Content as Inappropriate

Re: A query regarding inconsistency in Protege Ontology!

Michael DeBellis-2
Aparma, Glad we could help. If the reasoner doesn't give you errors then your ontology should be logically consistent. Good luck with your research.

Michael

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