Object Property Expression

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

Object Property Expression

Michael Meighu
Hi:

From my email yesterday - which I was asking what's the use case difference between the transitive property and the property chains axiom.  The property chains statement can be found at the Object Property level.  If you click on an object property, and you look in the description view, you hit the add icon next to the SuperProperyOf (Chain).  You then specify, "PropertyA o PropertyB".  That makes PropertyB a subclass of PropertyA.  

So let's say you have something that is made in Place A, and Place A is in Place B, then the ontology will entail that the something is also made in Place B.  

Maybe this is for individuals - while transitive is for classes....?

Mike 

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

Re: Object Property Expression

Lorenz B.
Did you carefully read all the answers?

There is a clear difference between transitive and a property chain used in a subclass axiom.
I means, the most obvious is that a property chain allows to "mix" different properties, or not?

And yes, of course you can just use a property chain axiom to express transitivity. r o r < r
You should know that property chains have been introduced in OWL 2 while transitivity could already be expressed in OWL 1.
Hi:

From my email yesterday - which I was asking what's the use case difference between the transitive property and the property chains axiom.  The property chains statement can be found at the Object Property level.  If you click on an object property, and you look in the description view, you hit the add icon next to the SuperProperyOf (Chain).  You then specify, "PropertyA o PropertyB".  That makes PropertyB a subclass of PropertyA.  

So let's say you have something that is made in Place A, and Place A is in Place B, then the ontology will entail that the something is also made in Place B.  

Maybe this is for individuals - while transitive is for classes....?

Mike 

_______________________________________________
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
Reply | Threaded
Open this post in threaded view
|

Re: Object Property Expression

Lorenz B.
In reply to this post by Michael Meighu



So let's say you have something that is made in Place A, and Place A is in Place B, then the ontology will entail that the something is also made in Place B.  

Maybe this is for individuals - while transitive is for classes....?


Why? You say a property is transitive, so why should it be "for classes"? You say the relation is transitive

No, both are used for relations between individuals - where exactly is the difficulty in understanding here?


-- 
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
Reply | Threaded
Open this post in threaded view
|

Re: Object Property Expression

Michael DeBellis-2
In reply to this post by Michael Meighu
Mike, I just want to clarify the terminology. There is a subtle difference between sub and super properties and subclasses. Where you said " That makes PropertyB a subclass of PropertyA." the correct terminology would be that  propertyB is a sub-property of propertyA not a subclass. They are very similar because properties are essentially equivalent to relations in logic and a relation is just a special kind of set. So set theoretically what you said is correct if you mean subset rather than subclass, a sub-property of propertyA is a subset of propertyA. But since a relation (aka property) is a set of ordered pairs it means something different to say that a property is a sub-property than it does to say a class is a subclass. I mention this because I find it confuses a lot of new users (it confused me when I was new) because it's one of the most significant ways that properties in OWL differ from properties in standard OOP. 

When B is a sub-property of A it means that all the ordered pairs in B are in A but not necessarily vice versa. A simple and obvious example of a sub-property would be the super-property hasParent which has sub-properties hasFather and hasMother. 

If  JohnSmith hasFather JoeSmith then the reasoner will infer that JohnSmith hasParent JoeSmith but if JohnSmith hasParent JoeSmith the reasoner won't infer (unless there is more information) that JohnSmith hasFather JoeSmith. Often for sub-properties the Domain and/or Range for the sub-property is a subclass of the Domain or Range of the super-property. So in the case of hasParent the domain and range could both be Person but the Range for hasFather would be Man (a subclass of Person) and the Range for hasMother would be Woman. 

And although Lorenz already said this, I just want to reiterate in both cases property chains and transitive properties it is about propagating values from one individual to another. Of course, the reasoner can always use information about properties as part of the definition and classification of the ontology so these definitions may impact how a specific class is classified but in terms of the use cases they are similar and always about propagating values from one individual to another. I almost never use property chains myself because anything you can do with a property chain you can do with a SWRL rule and I find SWRL rules more intuitive. But that's just my individual preference, not claiming it's the best way.

Michael

On Thu, Jan 3, 2019 at 5:40 AM Michael Meighu <[hidden email]> wrote:
Hi:

From my email yesterday - which I was asking what's the use case difference between the transitive property and the property chains axiom.  The property chains statement can be found at the Object Property level.  If you click on an object property, and you look in the description view, you hit the add icon next to the SuperProperyOf (Chain).  You then specify, "PropertyA o PropertyB".  That makes PropertyB a subclass of PropertyA.  

So let's say you have something that is made in Place A, and Place A is in Place B, then the ontology will entail that the something is also made in Place B.  

Maybe this is for individuals - while transitive is for classes....?

Mike 
_______________________________________________
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