Why isn't Pizza the Parent of CheesyPizza?

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

Why isn't Pizza the Parent of CheesyPizza?

kcason
Can anyone please explain to me why Pizza isn't specifically stated as the
Parent of CheesyPizza? This seems like an obvious relationship to me but it
isn't in the ontology.
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy1.jpg>


Having said that, CheesyPizza still shows up as a child of Pizza in the
Entity Graph:
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy2.jpg>




--
Sent from: http://protege-project.136.n4.nabble.com/Protege-User-f4659818.html
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: Why isn't Pizza the Parent of CheesyPizza?

Michael DeBellis-2
Good question. First, I note that there seems to be a warning message across the top of your screen in the right corner that says "Web Protege is not configured properly".  That might be the problem. 

Also, remember that there is no reasoner in Web Protege yet so things that you depend on the reasoner to do won't happen when you load your ontology into Web Protege, although I don't think that is causing whatever is going on with the Pizza ontology. The UI shows it as a subclass of Pizza so I think Pizza should be showing up as it's parent. 

Can you attach your ontology in a message. Then we can look at it and see if we can replicate the problem. 

Cheers,
Michael

On Fri, Apr 10, 2020 at 12:10 PM kcason <[hidden email]> wrote:
Can anyone please explain to me why Pizza isn't specifically stated as the
Parent of CheesyPizza? This seems like an obvious relationship to me but it
isn't in the ontology.
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy1.jpg>


Having said that, CheesyPizza still shows up as a child of Pizza in the
Entity Graph:
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy2.jpg>




--
Sent from: http://protege-project.136.n4.nabble.com/Protege-User-f4659818.html
_______________________________________________
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: Why isn't Pizza the Parent of CheesyPizza?

samsontu
In reply to this post by kcason
Hi,

On Apr 10, 2020, at 12:09 PM, kcason <[hidden email]> wrote:

Can anyone please explain to me why Pizza isn't specifically stated as the
Parent of CheesyPizza? This seems like an obvious relationship to me but it
isn't in the ontology.
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy1.jpg>


Having said that, CheesyPizza still shows up as a child of Pizza in the
Entity Graph:
<http://protege-project.136.n4.nabble.com/file/t376816/cheesy2.jpg>


That’s an interesting question. The reason that Pizza is not specifically shown as the superclass of CheesyPizza, I think, is because the specific superclass/subclass relationship is not asserted in the ontology. Instead, the asserted axiom is that CheesyPizza is equivalent to (Pizza and (hasTopping some CheeseTopping)), as shown in the OWL Entity Description Editor below. Nevertheless, WebProtege shows the (hasTopping CheeseTopping) relation and Pizza is shown as a parent in the entity graph, so WebProtege’s behavior is inconsistent. I’ve filed a WebProtege issue (#672) on GitHub for this.

With best regards,
Samson



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

Re: Why isn't Pizza the Parent of CheesyPizza?

kcason
In reply to this post by Michael DeBellis-2
pizza.owl <http://protege-project.136.n4.nabble.com/file/t376816/pizza.owl>  
Here is the pizza.owl file I am using.



--
Sent from: http://protege-project.136.n4.nabble.com/Protege-User-f4659818.html
_______________________________________________
protege-user mailing list
[hidden email]
https://mailman.stanford.edu/mailman/listinfo/protege-user
Reply | Threaded
Open this post in threaded view
|

Re: Why isn't Pizza the Parent of CheesyPizza?

Michael DeBellis-2
I think Samson already addressed that this is a bug with Web Protege. But I tried an experiment just to check. It is the difference between defined classes (classes with necessary and sufficient conditions, the ones with three horizontal bars in the UI) and normal classes and the fact that Web Protege doesn't have a reasoner. 

When I loaded your ontology into Web Protege all the defined classes (at least all the ones I bothered to look at) had the same issue, their parents weren't displayed. However, when I ran the reasoner on your ontology in the Desktop version and then did Export Inferred Axioms and then loaded that ontology into Web Protege then those classes did have their parents displayed. It's why I think it is always a good idea to first do the export inferences option and load that export file into Web Protege or if you are developing in Web Protege to regularly take the ontology and bring it into the desktop version and run the reasoner. If you aren't familiar with the Export Inferred Axioms here is a blog post I wrote about it: https://www.michaeldebellis.com/post/export-inferred-axioms

Cheers,
Michael

On Mon, Apr 13, 2020 at 7:55 AM kcason <[hidden email]> wrote:
pizza.owl <http://protege-project.136.n4.nabble.com/file/t376816/pizza.owl
Here is the pizza.owl file I am using.



--
Sent from: http://protege-project.136.n4.nabble.com/Protege-User-f4659818.html
_______________________________________________
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: Why isn't Pizza the Parent of CheesyPizza?

samsontu
Michael,

I don’t think it’s a difference between defined classes and sufficient conditions. It’s a difference between asserted superclass being a class or a part of a class expression. If I assert that CheesyPizza is a subclass of (IceCream and (hasTopping some CheeseTopping)), WebProtege doesn’t show IceCream as a Parents of CheesyPizza in the Details view of CheesyPizza, but shows IceCream as a superclass in the Class Hierarchy and Entity Diagram views. 

I am not sure that I’ll call this a bug, as Pizza being a superclass in this case is an inferred axiom and WebProtege doesn’t promise inference. Nevertheless it’s definitely a case of WebProtege exhibiting inconsistent behaviors in its GUI.

With best regards,
Samson


On Apr 13, 2020, at 8:23 AM, Michael DeBellis <[hidden email]> wrote:

I think Samson already addressed that this is a bug with Web Protege. But I tried an experiment just to check. It is the difference between defined classes (classes with necessary and sufficient conditions, the ones with three horizontal bars in the UI) and normal classes and the fact that Web Protege doesn't have a reasoner. 

When I loaded your ontology into Web Protege all the defined classes (at least all the ones I bothered to look at) had the same issue, their parents weren't displayed. However, when I ran the reasoner on your ontology in the Desktop version and then did Export Inferred Axioms and then loaded that ontology into Web Protege then those classes did have their parents displayed. It's why I think it is always a good idea to first do the export inferences option and load that export file into Web Protege or if you are developing in Web Protege to regularly take the ontology and bring it into the desktop version and run the reasoner. If you aren't familiar with the Export Inferred Axioms here is a blog post I wrote about it: https://www.michaeldebellis.com/post/export-inferred-axioms

Cheers,
Michael

On Mon, Apr 13, 2020 at 7:55 AM kcason <[hidden email]> wrote:
pizza.owl <http://protege-project.136.n4.nabble.com/file/t376816/pizza.owl
Here is the pizza.owl file I am using.



--
Sent from: http://protege-project.136.n4.nabble.com/Protege-User-f4659818.html
_______________________________________________
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