[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

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

[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

Hai Wang
Hi Helen,
Protege is an OWL editor, and there is no problem for you to edit OWL-
FULL ontology. You just can not reasoning the OWL-FULL ontology you  
created.
To work on OWL FULL,  you need
* change the language profile to owl full: OWL->Preference->General -
 > choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.

Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.

Regards
Hai

On 6/19/06, [hidden email] <[hidden email]> wrote:
 >
 > Hai
 >
 > Thanks for your help.
 >
 > If for some "strange" reasons we stay with the OWL-FULL ontology,  
am I right
 > to conclude that protege owl plug-in is not the tool to author such
 > ontology?
 >
 > My experience is this: although when starting a new project, you  
are asked
 > to select the flavour of owl (Lite, DL, Full), even if we select  
OWL FULL,
 > the protege editor does not allow one to enter the OWL-FULL type of
 > relationship, such as stated in my example:
 >
 > [
 >
 >  ProcedureA  hasExpectedOutcome  OutcomeA.
 >
 >  where ProcedureA and OutcomeA are classes, hasExpectedOutcome  is a
 > rdf:Property.
 > ]
 >
 > I was hoping that someone can show me how to enter such OWL-FULL
 > relationship using Protege owl plug-in.  Otherwise, I will have to  
go back
 > to the text editor, which I definitely would rather not to.
 >
 > Kind regards.
 >
 > Helen
 >
 >
 >
 >
 >
 >
 >
 >
 >
 >
 >
 > Hai Wang <[hidden email]>
 > Sent by: [hidden email]
 >
 >
 > 06/19/2006 06:24 AM
 >
 >
 > Please respond to
 >  [hidden email]
 >
 >
 > To [hidden email]
 >
 > cc
 >
 > Subject [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
 > express owl:sameAs in Protege
 >
 >
 >
 >
 >
 >
 > Hi Helen,
 >  Thanks for your clarification.  Now I understood your problem.  If
 >  you did not bother OWL Full, maybe you could try to define some meta
 >  level classes to  propagate the retractions to the class and its
 >  subclasses.
 >  For example, we define a class :Meta
 >
 >
 >  :B     rdfs:subClassOf  :A
 >  :B rdf:typeof :Meta
 >  :A rdf:typeof :Meta
 >
 >  Then you can put the restrictions to Meta. Meta-class is a common
 >  used way to create class templates in Protege-Frames model.
 >
 >  Regards
 >  Hai
 >
 >  On 15 Jun 2006, at 22:13, [hidden email] wrote:
 >
 >  >
 >  > John and Hai
 >  >
 >  > I apologize for my non-standard OWL syntax.
 >  >
 >  > I am used to write triples in N3.  So when I said:
 >  >
 >  > :A :somePropertyP :C,
 >  >
 >  > It is a simple triple
 >  >
 >  > A p C
 >  >
 >  > where A and C are classes.  OWL does not prohibit using such
 >  > triples.  But It make the ontology OWL Full
 >  >
 >  > Same N3 notations are used in my examples.
 >  >
 >  > Hai, you are right when you have  A -> p some C, you can make
 >  > inference as you stated.  But this is not what I was saying.
 >  >
 >  >
 >  > In my example,
 >  >
 >  > :Male  :include  :MaleAnt.
 >  > :John  rdf:typeOf :Male.
 >  >
 >  > here, "include" is just a property, which has Male class and
 >  > MaleAnt class as its subject and object respectively.  You can
 >  > further consider :include to be a inverse of rdfs:subClassOf  
such that
 >  >
 >  > :include  owl:inverseOf rdfs:subClassOf
 >  >
 >  > which means :include is a property defines relationship of
 >  > "superClassOf", then you will see no rdfs entailment rules nor owl
 >  > entailment rules will let you infer
 >  >
 >  > :john :include :MaleAnt.
 >  >
 >  >
 >  > One reason of getting into this "messy" OWL-FULL business:   we  
are
 >  > trying to model medical guideline and protocols in OWL.  We have
 >  > decided all concepts should be modelled as classes, not instances.
 >  > In some cases, we must say something like:
 >  >
 >  > ProcedureA  hasExpectedOutcome  OutcomeA.
 >  >
 >  > where ProcedureA and OutcomeA are classes, hasExpectedOutcome  
is a
 >  > rdf:Property.
 >  >
 >  >
 >  > Helen
 >  >
 >  >
 >  >
 >  >
 >  >
 >  > "John Goodwin" <[hidden email]>
 >  > Sent by: [hidden email]
 >  > 06/15/2006 03:56 PM
 >  > Please respond to
 >  > [hidden email]
 >  >
 >  >
 >  > To
 >  > <[hidden email]>
 >  > cc
 >  > Subject
 >  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
 >  > express owl:sameAs in Protege
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  > Just to clarify when you say
 >  >
 >  > :A :somePropP :C
 >  >
 >  > do you mean
 >  >
 >  > A -> p some C (in Manchester OWL syntax)
 >  >
 >  > or A p C where A and C are treated as instances and we our now in
 >  > OWL full?
 >  >
 >  > John
 >  >
 >  > ________________________________
 >  >
 >  > From: [hidden email] on behalf
 > of
 >  > [hidden email]
 >  > Sent: Thu 15/06/2006 18:50
 >  > To: [hidden email]
 >  > Cc: [hidden email]
 >  > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
 >  > editing and express owl:sameAs in Protege
 >  >
 >  >
 >  >
 >  > Hai
 >  >
 >  > One common mistake one can easily make is to think a property
 >  > assigned to a class must apply to its subclasses and every of its
 >  > members, as follows:
 >  >
 >  > Given
 >  > :A  :somePropertyP :C.  (class A has property P with values from
 >  > class C)
 >  > :B     rdfs:subClassOf  :A  (class B is a subClass of class A)
 >  > :x  rdf:type :A  (a is a member of class A)
 >  >
 >  > then one can get:
 >  >
 >  > :B  :somePropertyP :C.  (class B also  has property P with values
 >  > from class C) - this is wrong
 >  > :x  :somePropertyP :C. (member x has property P with values from
 >  > class C) - this is wrong too.
 >  >
 >  > Jos gave a real life example as following:
 >  >
 >  > :Male  :include  :MaleAnt.
 >  > :John  rdf:typeOf :Male.
 >  >
 >  > it would be ridiculous to infer that
 >  >
 >  > :John :include :MaleAnt.
 >  >
 >  >
 >  > Therefore, even though :A and :B are sub-classes of each other,  
the
 >  > properties assigned to each of them can not be passed along.  
Hence:
 >  >
 >  > {:A  :somePropertyP :C.
 >  > :B :somePropertyQ :D.
 >  > :A owl:equivalentClass :B.}
 >  >
 >  > an owl reasoner can not entail
 >  >
 >  > {:A :somePropertyQ :D.
 >  > :B :somePropertyP :C}.
 >  >
 >  > The entailment rules in RDFS are quite simple:
 >  >
 >  > http://www.w3.org/TR/rdf-mt/#RDFSRules
 >  >
 >  > It is quite different from OO paradigm.  I got confused many times
 >  > - still so from time to time :)
 >  >
 >  >
 >  > In you email, were you saying:
 >  >
 >  > Given the following triples:
 >  >
 >  > {:A  :somePropertyP :C.
 >  > :B :somePropertyQ :D.
 >  > :A owl:equivalentClass :B.
 >  > :F  :somePropertyP :C
 >  > }
 >  > now you can entail:
 >  >
 >  > {:A rdfs:subClassOf :F.
 >  > :B rdfs:subClassOf :F}.
 >  >
 >  > I found it hard to understand what rules protege uses to get this
 >  > inference.   There is no way that :F should be inferred as
 >  > superclass of :A.  They could every be disjoint classes.
 >  >
 >  >
 >  > Helen.
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  > Hai Wang <[hidden email]>
 >  > Sent by: [hidden email]
 >  >
 >  > 06/15/2006 06:55 AM
 >  > Please respond to
 >  > [hidden email]
 >  >
 >  >
 >  > To
 >  > [hidden email]
 >  > cc
 >  > Subject
 >  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
 >  > express owl:sameAs in Protege
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  >
 >  > Hi Helen,
 >  > I totally agree with what you said about the difference between
 >  > owl:equivalentClass and owl:sameAs. However, one thing I am not so
 >  > sure is that you said:
 >  > *****
 >  > {:A  :somePropertyP :C.
 >  > :B :somePropertyQ :D.
 >  > :A owl:equivalentClass :B.}
 >  >
 >  > an owl reasoner can not entail
 >  >
 >  > {:A :somePropertyQ :D.
 >  > :B :somePropertyP :C}.
 >  > *******
 >  >
 >  > If I am right, since class A and B have same instances, the  
inference
 >  > could be made.
 >  > For example, the above model says that
 >  > 1. any instances of A will have a value for the property P from  
the
 >  > class C
 >  > 2. A and B have exactly same set of instances,
 >  > Therefore, we can same that any instances of B will have a  
value for
 >  > property P from the class C as well, which means that
 >  > {:B :somePropertyP :C}.
 >  >
 >  > If fact, if we created a class "F =  p some C", and run  
classifier.
 >  > Both A and B will be classified as the subclasses of F.
 >  >
 >  >
 >  > Regards
 >  > Hai
 >  >
 >  >
 >  >
 >  >
 >  > On 14 Jun 2006, at 20:56, [hidden email] wrote:
 >  >
 >  > >
 >  > > Hi, John
 >  > >
 >  > > Yes, with this triple:
 >  > >
 >  > > :A owl:equivalentClass :B
 >  > >
 >  > > You are only saying the extension of class :A and :B are the  
same,
 >  > > i.e. class A and class B have same members.  It is perfectly  
fine
 >  > > to say
 >  > >
 >  > > :A owl:equivalentClass :B.
 >  > > :A owl:differentFrom :B.
 >  > >
 >  > > This is because although :A and :B have the same membership,  
they
 >  > > are really two different concepts.
 >  > >
 >  > > Therefore, given
 >  > >
 >  > > {:A  :somePropertyP :C.
 >  > > :B :somePropertyQ :D.
 >  > > :A owl:equivalentClass :B.}
 >  > >
 >  > > an owl reasoner can not entail
 >  > >
 >  > > {:A :somePropertyQ :D.
 >  > > :B :somePropertyP :C}.
 >  > >
 >  > > But when you say:
 >  > >
 >  > > :A owl:sameAs :B.
 >  > >
 >  > > This means :A and :B are really the same thing (uri).   So if :A
 >  > > and :B are classes, all members of class A are also members of
 >  > > class B, and vice versa.
 >  > >
 >  > > Furthermore, given:
 >  > >
 >  > > {:A  :somePropertyP :C.
 >  > > :B :somePropertyQ :D.
 >  > > :A owl:sameAs :B.}
 >  > >
 >  > > an owl reason can entail:
 >  > >
 >  > > {:A :somePropertyQ :D.
 >  > > :B :somePropertyP :C}.
 >  > >
 >  > > I have just learnt the above difference from Jos De Roo today.
 >  > > Hope this helps :)
 >  > >
 >  > > Kind regards.
 >  > >
 >  > > Helen
 >  > >
 >  > >
 >  > >
 >  > >
 >  > > "John Goodwin" <[hidden email]>
 >  > > Sent by: [hidden email]
 >  > > 06/14/2006 10:49 AM
 >  > > Please respond to
 >  > > [hidden email]
 >  > >
 >  > >
 >  > > To
 >  > > <[hidden email]>
 >  > > cc
 >  > > Subject
 >  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
 >  > > express owl:sameAs in Protege
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > > You mean with equivalentTo you can only say the classes have the
 >  > > same extension? Yes that's true A equivalentTo B does not  
imply A
 >  > > is the same concept as B. However, as far as machine  
reasoning goes
 >  > > isn't equivalentTo sufficient? It's perhaps more of user  
interface
 >  > > issue to state they are the same concept. Could you add some
 >  > > property "sameConceptAs"? So you could say:
 >  > >
 >  > > A equivalentTo B
 >  > > A -> sameConceptAs some B
 >  > > B -> sameConceptAs some A
 >  > > sameConceptAs is transitive and symmetric
 >  > >
 >  > > Would that then capture the require machine semantics, and  
also the
 >  > > knowledge they are the same concept? Or is that a bit messy?
 >  > >
 >  > > I've never used Euler or CWM...but aren't the semantics for OWL
 >  > > Full hard to define? Might be ok in this case. I know OWL 1.1
 >  > > allows this sort of punning. To quote:
 >  > >
 >  > > In OWL 1.1 a name (like Person) can be used as both an  
individual
 >  > > and a class and a property. The computational problems that  
would
 >  > > arise if this was treated as in RDF are avoided by ensuring  
that no
 >  > > aspect of the use of the name as an individual has any effect on
 >  > > the meaning of the name as a class. (This treatment is often  
called
 >  > > punning.)
 >  > >
 >  > > With the above change, general properties can be placed on names
 >  > > that are used as classes, and there is additional syntactic  
sugar
 >  > > to allow the placement of property values in class and property
 >  > > axioms, as in Class(Person partial super(Animal) type
(MyClass) value
 >  > > (dc:creator peter)) which is syntactic sugar for Class(Person
 >  > > partial) plus Individual(Person super(Animal) type(MyClass)  
value
 >  > > (dc:creator peter)). (The above example assumes a definition for
 >  > > dc:creator as an individual-valued property.)
 >  > >
 >  > >
 >  > > From
 > http://www-db.research.bell-labs.com/user/pfps/owl/
 >  > overview.html
 >  > >
 >  > > I've never really ventured into OWL full so can't really help  
too
 >  > > much I'm afraid.
 >  > >
 >  > > John
 >  > >
 >  > >
 >  > > From: [hidden email]
 > [mailto:protege-owl-
 >  > > [hidden email]] On Behalf Of [hidden email]
 >  > > Sent: 14 June 2006 15:15
 >  > > To: [hidden email]
 >  > > Cc: [hidden email]; protege-owl-bounce@crg-
 >  > > gw.Stanford.EDU
 >  > > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
 >  > > editing and express owl:sameAs in Protege
 >  > >
 >  > >
 >  > > Raj and John
 >  > >
 >  > > Thanks for your reply.  My experience is exactly what Raj
 >  > > described, word for word.
 >  > >
 >  > > John, here is the reason I want to use owl:equivalentClass:
 >  > >
 >  > > When trying to map two knowledge bases, I have a
 >  > > concept  :GestationalPatient in OntologyA,  which is defined  
as any
 >  > > patient from 0 week to the moment of birth.  In OntologyB, I  
have a
 >  > > concept :InUteroPatient, which defined as any patient in  
ulterus.
 >  > > When talking about human patients, these two concepts are really
 >  > > the same concept, and this is what I want to express with
 >  > > owl:equivalentClass.
 >  > >
 >  > > I can either write
 >  > >
 >  > > :GestationalPatient owl:sameAs :InUteroPatient
 >  > >
 >  > > or
 >  > >
 >  > > :GestationalPatient owl:equivalentClass :InUteroPatient
 >  > >
 >  > > But with the owl:sameAs relationship, I can only say these two
 >  > > classes have the same extension, ie, exactly the same  
members.  I
 >  > > really want to say these two concepts are the same concept, not
 >  > > just happend to have the same members.
 >  > >
 >  > > We use semantic web reasoners such as euler and cwm, so being  
OWL-
 >  > > FULL is not a problem for me.
 >  > >
 >  > > I read the protege-owl tutorial and was expecting to see  
those owl
 >  > > properties displayed when I checked them all on performance-
 >  > > >visibility menu, but didn't see anything.
 >  > >
 >  > > Kind regards.
 >  > >
 >  > > Helen
 >  > >
 >  > >
 >  > >
 >  > > "John Goodwin" <[hidden email]>
 >  > > Sent by: [hidden email]
 >  > > 06/14/2006 07:56 AM
 >  > > Please respond to
 >  > > [hidden email]
 >  > >
 >  > >
 >  > >
 >  > > To
 >  > > <[hidden email]>
 >  > > cc
 >  > > Subject
 >  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
 >  > > express owl:sameAs in Protege
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > >
 >  > > Answer below...
 >  > >
 >  > > I read these sentences, of your previous response, without  
giving a
 >  > > break in between... I didn't see that these sentences were  
answers
 >  > > for two different questions... so I was expecting to see this
 >  > > 'equivalentTo' display in the class hierarchy after going to the
 >  > > OWL menu and select preferences ;)
 >  > >
 >  > > Ahh sorry for the confusion!
 >  > >
 >  > > What I meant with with OWL Properties is, when you go to the OWL
 >  > > menu, then select Preferences, and then select the Visibility  
tab,
 >  > > and here when you can see all of the Annotation Properties
 >  > > selected, but none of the OWL Properties are checked (atleast  
for
 >  > > me) so what is supposed to happen when we check and uncheck  
these
 >  > > properties? there is something to do with visibilitiy, so I
 >  > > expected that they show a widget (pane, or panel) or  
something like
 >  > > that for every kind of property... but the only widget shown is
 >  > > for 'owl:disjointWith' in the class editor window... but I
 >  > > don't know how different is this 'owl:disjointWith' widget
 >  > > from the normal Disjoints widget b'cos both are showing the same
 >  > > values...
 >  > >
 >  > >
 >  > > I've not really played with or used these preferences so not  
really
 >  > > sure I'm afraid.
 >  > >
 >  > > John
 >  > > .
 >  > >
 >  > >
 >  > > This email is only intended for the person to whom it is  
addressed
 >  > > and may contain confidential information. If you have  
received this
 >  > > email in error, please notify the sender and delete this email
 >  > > which must not be copied, distributed or disclosed to any other
 >  > > person.
 >  > >
 >  > > Unless stated otherwise, the contents of this email are  
personal to
 >  > > the writer and do not represent the official view of Ordnance
 >  > > Survey. Nor can any contract be formed on Ordnance Survey's  
behalf
 >  > > via email. We reserve the right to monitor emails and  
attachments
 >  > > without prior notice.
 >  > >
 >  > > Thank you for your cooperation.
 >  > >
 >  > > Ordnance Survey
 >  > > Romsey Road
 >  > > Southampton SO16 4GU
 >  > > Tel: 023 8079 2000
 >  > > http://www.ordnancesurvey.co.uk
 >  > >
 >  > > .
 >  > >
 >  > >
 >  > > This email is only intended for the person to whom it is  
addressed
 >  > > and may contain confidential information. If you have  
received this
 >  > > email in error, please notify the sender and delete this email
 >  > > which must not be copied, distributed or disclosed to any other
 >  > > person.
 >  > >
 >  > > Unless stated otherwise, the contents of this email are  
personal to
 >  > > the writer and do not represent the official view of Ordnance
 >  > > Survey. Nor can any contract be formed on Ordnance Survey's  
behalf
 >  > > via email. We reserve the right to monitor emails and  
attachments
 >  > > without prior notice.
 >  > >
 >  > > Thank you for your cooperation.
 >  > >
 >  > > Ordnance Survey
 >  > > Romsey Road
 >  > > Southampton SO16 4GU
 >  > > Tel: 023 8079 2000
 >  > > http://www.ordnancesurvey.co.uk
 >  > >
 >  > >
 >  > >
 >  >
 >  >
 >  
----------------------------------------------------------------------
 >  > ---
 >  > To unsubscribe go to
 > http://protege.stanford.edu/community/
 >  > subscribe.html
 >  >
 >  >
 >  >
 >  >
 >  > .
 >  >
 >  >
 >  > This email is only intended for the person to whom it is addressed
 >  > and may contain confidential information. If you have received  
this
 >  > email in error, please notify the sender and delete this email
 >  > which must not be copied, distributed or disclosed to any other
 >  > person.
 >  >
 >  > Unless stated otherwise, the contents of this email are  
personal to
 >  > the writer and do not represent the official view of Ordnance
 >  > Survey. Nor can any contract be formed on Ordnance Survey's behalf
 >  > via email. We reserve the right to monitor emails and attachments
 >  > without prior notice.
 >  >
 >  > Thank you for your cooperation.
 >  >
 >  > Ordnance Survey
 >  > Romsey Road
 >  > Southampton SO16 4GU
 >  > Tel: 023 8079 2000
 >  > http://www.ordnancesurvey.co.uk
 >  >
 >  >
 >  > <winmail.dat>
 >
 >  
------------------------------------------------------------------------
-
 >  To unsubscribe go to
 > http://protege.stanford.edu/community/subscribe.html
 >
 >
 >


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

helen.chen

Hai

Thanks so much.  This is what I was looking for.  But I still need some coaching,

[* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.
]

yes, I did this and now I can see classes in the "individual tab". However, it seems I must declare my classes are all subclass of owl:Class in order to be treated as instances.

[Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.
]

This part I didn't get it.

For example, I have class A and B, and I want to say

:A  owl:sameAs :B,

Now I made them subclasses of owl:Class, so I can see them in individual tab.

I also made owl:sameAs property visible.  So I can also see them in as instances of rdf:Property in the individual tab.

What steps I shall follow in order to declare the triple:

:A  owl:sameAs :B


Furthermore, if I want to say a simple triple

:a  :p  :b

where :a is an instance of class A and :b is an instance of class B, :p is an owl:ObjectProperty, without declaring domain and range on property p, I could not find property p in the individual editor.

I am just one step away but could not find the path to go there :)


Thanks  

Helen




Hai Wang <[hidden email]>
Sent by: [hidden email]

06/20/2006 05:21 AM

Please respond to
[hidden email]

To
[hidden email]
cc
Subject
[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege






Hi Helen,
Protege is an OWL editor, and there is no problem for you to edit OWL-
FULL ontology. You just can not reasoning the OWL-FULL ontology you  
created.
To work on OWL FULL,  you need
* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.

Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.

Regards
Hai

On 6/19/06, [hidden email] <[hidden email]> wrote:
>
> Hai
>
> Thanks for your help.
>
> If for some "strange" reasons we stay with the OWL-FULL ontology,  
am I right
> to conclude that protege owl plug-in is not the tool to author such
> ontology?
>
> My experience is this: although when starting a new project, you  
are asked
> to select the flavour of owl (Lite, DL, Full), even if we select  
OWL FULL,
> the protege editor does not allow one to enter the OWL-FULL type of
> relationship, such as stated in my example:
>
> [
>
>  ProcedureA  hasExpectedOutcome  OutcomeA.
>
>  where ProcedureA and OutcomeA are classes, hasExpectedOutcome  is a
> rdf:Property.
> ]
>
> I was hoping that someone can show me how to enter such OWL-FULL
> relationship using Protege owl plug-in.  Otherwise, I will have to  
go back
> to the text editor, which I definitely would rather not to.
>
> Kind regards.
>
> Helen
>
>
>
>
>
>
>
>
>
>
>
> Hai Wang <[hidden email]>
> Sent by: [hidden email]
>
>
> 06/19/2006 06:24 AM
>
>
> Please respond to
>  [hidden email]
>
>
> To [hidden email]
>
> cc
>
> Subject [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
> express owl:sameAs in Protege
>
>
>
>
>
>
> Hi Helen,
>  Thanks for your clarification.  Now I understood your problem.  If
>  you did not bother OWL Full, maybe you could try to define some meta
>  level classes to  propagate the retractions to the class and its
>  subclasses.
>  For example, we define a class :Meta
>
>
>  :B     rdfs:subClassOf  :A
>  :B rdf:typeof :Meta
>  :A rdf:typeof :Meta
>
>  Then you can put the restrictions to Meta. Meta-class is a common
>  used way to create class templates in Protege-Frames model.
>
>  Regards
>  Hai
>
>  On 15 Jun 2006, at 22:13, [hidden email] wrote:
>
>  >
>  > John and Hai
>  >
>  > I apologize for my non-standard OWL syntax.
>  >
>  > I am used to write triples in N3.  So when I said:
>  >
>  > :A :somePropertyP :C,
>  >
>  > It is a simple triple
>  >
>  > A p C
>  >
>  > where A and C are classes.  OWL does not prohibit using such
>  > triples.  But It make the ontology OWL Full
>  >
>  > Same N3 notations are used in my examples.
>  >
>  > Hai, you are right when you have  A -> p some C, you can make
>  > inference as you stated.  But this is not what I was saying.
>  >
>  >
>  > In my example,
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > here, "include" is just a property, which has Male class and
>  > MaleAnt class as its subject and object respectively.  You can
>  > further consider :include to be a inverse of rdfs:subClassOf  
such that
>  >
>  > :include  owl:inverseOf rdfs:subClassOf
>  >
>  > which means :include is a property defines relationship of
>  > "superClassOf", then you will see no rdfs entailment rules nor owl
>  > entailment rules will let you infer
>  >
>  > :john :include :MaleAnt.
>  >
>  >
>  > One reason of getting into this "messy" OWL-FULL business:   we  
are
>  > trying to model medical guideline and protocols in OWL.  We have
>  > decided all concepts should be modelled as classes, not instances.
>  > In some cases, we must say something like:
>  >
>  > ProcedureA  hasExpectedOutcome  OutcomeA.
>  >
>  > where ProcedureA and OutcomeA are classes, hasExpectedOutcome  
is a
>  > rdf:Property.
>  >
>  >
>  > Helen
>  >
>  >
>  >
>  >
>  >
>  > "John Goodwin" <[hidden email]>
>  > Sent by: [hidden email]
>  > 06/15/2006 03:56 PM
>  > Please respond to
>  > [hidden email]
>  >
>  >
>  > To
>  > <[hidden email]>
>  > cc
>  > Subject
>  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
>  > express owl:sameAs in Protege
>  >
>  >
>  >
>  >
>  >
>  >
>  > Just to clarify when you say
>  >
>  > :A :somePropP :C
>  >
>  > do you mean
>  >
>  > A -> p some C (in Manchester OWL syntax)
>  >
>  > or A p C where A and C are treated as instances and we our now in
>  > OWL full?
>  >
>  > John
>  >
>  > ________________________________
>  >
>  > From: [hidden email] on behalf
> of
>  > [hidden email]
>  > Sent: Thu 15/06/2006 18:50
>  > To: [hidden email]
>  > Cc: [hidden email]
>  > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
>  > editing and express owl:sameAs in Protege
>  >
>  >
>  >
>  > Hai
>  >
>  > One common mistake one can easily make is to think a property
>  > assigned to a class must apply to its subclasses and every of its
>  > members, as follows:
>  >
>  > Given
>  > :A  :somePropertyP :C.  (class A has property P with values from
>  > class C)
>  > :B     rdfs:subClassOf  :A  (class B is a subClass of class A)
>  > :x  rdf:type :A  (a is a member of class A)
>  >
>  > then one can get:
>  >
>  > :B  :somePropertyP :C.  (class B also  has property P with values
>  > from class C) - this is wrong
>  > :x  :somePropertyP :C. (member x has property P with values from
>  > class C) - this is wrong too.
>  >
>  > Jos gave a real life example as following:
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > it would be ridiculous to infer that
>  >
>  > :John :include :MaleAnt.
>  >
>  >
>  > Therefore, even though :A and :B are sub-classes of each other,  
the
>  > properties assigned to each of them can not be passed along.  
Hence:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.}
>  >
>  > an owl reasoner can not entail
>  >
>  > {:A :somePropertyQ :D.
>  > :B :somePropertyP :C}.
>  >
>  > The entailment rules in RDFS are quite simple:
>  >
>  > http://www.w3.org/TR/rdf-mt/#RDFSRules
>  >
>  > It is quite different from OO paradigm.  I got confused many times
>  > - still so from time to time :)
>  >
>  >
>  > In you email, were you saying:
>  >
>  > Given the following triples:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.
>  > :F  :somePropertyP :C
>  > }
>  > now you can entail:
>  >
>  > {:A rdfs:subClassOf :F.
>  > :B rdfs:subClassOf :F}.
>  >
>  > I found it hard to understand what rules protege uses to get this
>  > inference.   There is no way that :F should be inferred as
>  > superclass of :A.  They could every be disjoint classes.
>  >
>  >
>  > Helen.
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  > Hai Wang <[hidden email]>
>  > Sent by: [hidden email]
>  >
>  > 06/15/2006 06:55 AM
>  > Please respond to
>  > [hidden email]
>  >
>  >
>  > To
>  > [hidden email]
>  > cc
>  > Subject
>  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
>  > express owl:sameAs in Protege
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  > Hi Helen,
>  > I totally agree with what you said about the difference between
>  > owl:equivalentClass and owl:sameAs. However, one thing I am not so
>  > sure is that you said:
>  > *****
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.}
>  >
>  > an owl reasoner can not entail
>  >
>  > {:A :somePropertyQ :D.
>  > :B :somePropertyP :C}.
>  > *******
>  >
>  > If I am right, since class A and B have same instances, the  
inference
>  > could be made.
>  > For example, the above model says that
>  > 1. any instances of A will have a value for the property P from  
the
>  > class C
>  > 2. A and B have exactly same set of instances,
>  > Therefore, we can same that any instances of B will have a  
value for
>  > property P from the class C as well, which means that
>  > {:B :somePropertyP :C}.
>  >
>  > If fact, if we created a class "F =  p some C", and run  
classifier.
>  > Both A and B will be classified as the subclasses of F.
>  >
>  >
>  > Regards
>  > Hai
>  >
>  >
>  >
>  >
>  > On 14 Jun 2006, at 20:56, [hidden email] wrote:
>  >
>  > >
>  > > Hi, John
>  > >
>  > > Yes, with this triple:
>  > >
>  > > :A owl:equivalentClass :B
>  > >
>  > > You are only saying the extension of class :A and :B are the  
same,
>  > > i.e. class A and class B have same members.  It is perfectly  
fine
>  > > to say
>  > >
>  > > :A owl:equivalentClass :B.
>  > > :A owl:differentFrom :B.
>  > >
>  > > This is because although :A and :B have the same membership,  
they
>  > > are really two different concepts.
>  > >
>  > > Therefore, given
>  > >
>  > > {:A  :somePropertyP :C.
>  > > :B :somePropertyQ :D.
>  > > :A owl:equivalentClass :B.}
>  > >
>  > > an owl reasoner can not entail
>  > >
>  > > {:A :somePropertyQ :D.
>  > > :B :somePropertyP :C}.
>  > >
>  > > But when you say:
>  > >
>  > > :A owl:sameAs :B.
>  > >
>  > > This means :A and :B are really the same thing (uri).   So if :A
>  > > and :B are classes, all members of class A are also members of
>  > > class B, and vice versa.
>  > >
>  > > Furthermore, given:
>  > >
>  > > {:A  :somePropertyP :C.
>  > > :B :somePropertyQ :D.
>  > > :A owl:sameAs :B.}
>  > >
>  > > an owl reason can entail:
>  > >
>  > > {:A :somePropertyQ :D.
>  > > :B :somePropertyP :C}.
>  > >
>  > > I have just learnt the above difference from Jos De Roo today.
>  > > Hope this helps :)
>  > >
>  > > Kind regards.
>  > >
>  > > Helen
>  > >
>  > >
>  > >
>  > >
>  > > "John Goodwin" <[hidden email]>
>  > > Sent by: [hidden email]
>  > > 06/14/2006 10:49 AM
>  > > Please respond to
>  > > [hidden email]
>  > >
>  > >
>  > > To
>  > > <[hidden email]>
>  > > cc
>  > > Subject
>  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
>  > > express owl:sameAs in Protege
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > > You mean with equivalentTo you can only say the classes have the
>  > > same extension? Yes that's true A equivalentTo B does not  
imply A
>  > > is the same concept as B. However, as far as machine  
reasoning goes
>  > > isn't equivalentTo sufficient? It's perhaps more of user  
interface
>  > > issue to state they are the same concept. Could you add some
>  > > property "sameConceptAs"? So you could say:
>  > >
>  > > A equivalentTo B
>  > > A -> sameConceptAs some B
>  > > B -> sameConceptAs some A
>  > > sameConceptAs is transitive and symmetric
>  > >
>  > > Would that then capture the require machine semantics, and  
also the
>  > > knowledge they are the same concept? Or is that a bit messy?
>  > >
>  > > I've never used Euler or CWM...but aren't the semantics for OWL
>  > > Full hard to define? Might be ok in this case. I know OWL 1.1
>  > > allows this sort of punning. To quote:
>  > >
>  > > In OWL 1.1 a name (like Person) can be used as both an  
individual
>  > > and a class and a property. The computational problems that  
would
>  > > arise if this was treated as in RDF are avoided by ensuring  
that no
>  > > aspect of the use of the name as an individual has any effect on
>  > > the meaning of the name as a class. (This treatment is often  
called
>  > > punning.)
>  > >
>  > > With the above change, general properties can be placed on names
>  > > that are used as classes, and there is additional syntactic  
sugar
>  > > to allow the placement of property values in class and property
>  > > axioms, as in Class(Person partial super(Animal) type
(MyClass) value
>  > > (dc:creator peter)) which is syntactic sugar for Class(Person
>  > > partial) plus Individual(Person super(Animal) type(MyClass)  
value
>  > > (dc:creator peter)). (The above example assumes a definition for
>  > > dc:creator as an individual-valued property.)
>  > >
>  > >
>  > > From
> http://www-db.research.bell-labs.com/user/pfps/owl/
>  > overview.html
>  > >
>  > > I've never really ventured into OWL full so can't really help  
too
>  > > much I'm afraid.
>  > >
>  > > John
>  > >
>  > >
>  > > From: [hidden email]
> [mailto:protege-owl-
>  > > [hidden email]] On Behalf Of [hidden email]
>  > > Sent: 14 June 2006 15:15
>  > > To: [hidden email]
>  > > Cc: [hidden email]; protege-owl-bounce@crg-
>  > > gw.Stanford.EDU
>  > > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
>  > > editing and express owl:sameAs in Protege
>  > >
>  > >
>  > > Raj and John
>  > >
>  > > Thanks for your reply.  My experience is exactly what Raj
>  > > described, word for word.
>  > >
>  > > John, here is the reason I want to use owl:equivalentClass:
>  > >
>  > > When trying to map two knowledge bases, I have a
>  > > concept  :GestationalPatient in OntologyA,  which is defined  
as any
>  > > patient from 0 week to the moment of birth.  In OntologyB, I  
have a
>  > > concept :InUteroPatient, which defined as any patient in  
ulterus.
>  > > When talking about human patients, these two concepts are really
>  > > the same concept, and this is what I want to express with
>  > > owl:equivalentClass.
>  > >
>  > > I can either write
>  > >
>  > > :GestationalPatient owl:sameAs :InUteroPatient
>  > >
>  > > or
>  > >
>  > > :GestationalPatient owl:equivalentClass :InUteroPatient
>  > >
>  > > But with the owl:sameAs relationship, I can only say these two
>  > > classes have the same extension, ie, exactly the same  
members.  I
>  > > really want to say these two concepts are the same concept, not
>  > > just happend to have the same members.
>  > >
>  > > We use semantic web reasoners such as euler and cwm, so being  
OWL-
>  > > FULL is not a problem for me.
>  > >
>  > > I read the protege-owl tutorial and was expecting to see  
those owl
>  > > properties displayed when I checked them all on performance-
>  > > >visibility menu, but didn't see anything.
>  > >
>  > > Kind regards.
>  > >
>  > > Helen
>  > >
>  > >
>  > >
>  > > "John Goodwin" <[hidden email]>
>  > > Sent by: [hidden email]
>  > > 06/14/2006 07:56 AM
>  > > Please respond to
>  > > [hidden email]
>  > >
>  > >
>  > >
>  > > To
>  > > <[hidden email]>
>  > > cc
>  > > Subject
>  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
>  > > express owl:sameAs in Protege
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > > Answer below...
>  > >
>  > > I read these sentences, of your previous response, without  
giving a
>  > > break in between... I didn't see that these sentences were  
answers
>  > > for two different questions... so I was expecting to see this
>  > > 'equivalentTo' display in the class hierarchy after going to the
>  > > OWL menu and select preferences ;)
>  > >
>  > > Ahh sorry for the confusion!
>  > >
>  > > What I meant with with OWL Properties is, when you go to the OWL
>  > > menu, then select Preferences, and then select the Visibility  
tab,
>  > > and here when you can see all of the Annotation Properties
>  > > selected, but none of the OWL Properties are checked (atleast  
for
>  > > me) so what is supposed to happen when we check and uncheck  
these
>  > > properties? there is something to do with visibilitiy, so I
>  > > expected that they show a widget (pane, or panel) or  
something like
>  > > that for every kind of property... but the only widget shown is
>  > > for 'owl:disjointWith' in the class editor window... but I

 >  > > don't know how different is this 'owl:disjointWith' widget
>  > > from the normal Disjoints widget b'cos both are showing the same
>  > > values...
>  > >
>  > >
>  > > I've not really played with or used these preferences so not  
really
>  > > sure I'm afraid.
>  > >
>  > > John
>  > > .
>  > >
>  > >
>  > > This email is only intended for the person to whom it is  
addressed
>  > > and may contain confidential information. If you have  
received this
>  > > email in error, please notify the sender and delete this email
>  > > which must not be copied, distributed or disclosed to any other
>  > > person.
>  > >
>  > > Unless stated otherwise, the contents of this email are  
personal to
>  > > the writer and do not represent the official view of Ordnance
>  > > Survey. Nor can any contract be formed on Ordnance Survey's  
behalf
>  > > via email. We reserve the right to monitor emails and  
attachments
>  > > without prior notice.
>  > >
>  > > Thank you for your cooperation.
>  > >
>  > > Ordnance Survey
>  > > Romsey Road
>  > > Southampton SO16 4GU
>  > > Tel: 023 8079 2000
>  > > http://www.ordnancesurvey.co.uk
>  > >
>  > > .
>  > >
>  > >
>  > > This email is only intended for the person to whom it is  
addressed
>  > > and may contain confidential information. If you have  
received this
>  > > email in error, please notify the sender and delete this email
>  > > which must not be copied, distributed or disclosed to any other
>  > > person.
>  > >
>  > > Unless stated otherwise, the contents of this email are  
personal to
>  > > the writer and do not represent the official view of Ordnance
>  > > Survey. Nor can any contract be formed on Ordnance Survey's  
behalf
>  > > via email. We reserve the right to monitor emails and  
attachments
>  > > without prior notice.
>  > >
>  > > Thank you for your cooperation.
>  > >
>  > > Ordnance Survey
>  > > Romsey Road
>  > > Southampton SO16 4GU
>  > > Tel: 023 8079 2000
>  > > http://www.ordnancesurvey.co.uk
>  > >
>  > >
>  > >
>  >
>  >
>  
----------------------------------------------------------------------
>  > ---
>  > To unsubscribe go to
> http://protege.stanford.edu/community/
>  > subscribe.html
>  >
>  >
>  >
>  >
>  > .
>  >
>  >
>  > This email is only intended for the person to whom it is addressed
>  > and may contain confidential information. If you have received  
this
>  > email in error, please notify the sender and delete this email
>  > which must not be copied, distributed or disclosed to any other
>  > person.
>  >
>  > Unless stated otherwise, the contents of this email are  
personal to
>  > the writer and do not represent the official view of Ordnance
>  > Survey. Nor can any contract be formed on Ordnance Survey's behalf
>  > via email. We reserve the right to monitor emails and attachments
>  > without prior notice.
>  >
>  > Thank you for your cooperation.
>  >
>  > Ordnance Survey
>  > Romsey Road
>  > Southampton SO16 4GU
>  > Tel: 023 8079 2000
>  > http://www.ordnancesurvey.co.uk
>  >
>  >
>  > <winmail.dat>
>
>  
------------------------------------------------------------------------
-
>  To unsubscribe go to
> http://protege.stanford.edu/community/subscribe.html
>
>
>


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html


Reply | Threaded
Open this post in threaded view
|

[protege-owl] AW: Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

Rajverma
In reply to this post by Hai Wang

Hi Hai,

 

when I did the first step, in the individual browser it says that there are 32 instances (asserted) of rdfs:Class and 11 instances (asserted) of owl:Class, however in the instance browser I can see only 18 instances (asserted) for rdfs:Class and 3 instances (asserted) for owl:Class… where are the rest?

 

Regarding owl:sameAs, I’m still waiting for some one to tell how to use it… I also made this property visible by checking the box, but I can’t see it as an instance of rdfProperty in the individual tab as Helen cud see…

 

Helen, regarding your last question, you can see what you want in the Forms tab… but I don’t know what we can do with this object property pane in the form editor b’cos I can’t select any thing into it as the buttons inside it are inactive, unlike what we have in Individuals editor…

 

If you want to see what you want in the Individuals editor then you must create individuals for the classes by clicking the purple diamond button with a + sign on the top in the instance browser… Once you have an asserted individual (in the instance browser) for your classes and if the class of this individual has a property then that property will be shown in the individual editor… is this what you want?

 

Have a look at the attachment…

 

Cheers,

Raj

 

 


Von: [hidden email] [mailto:[hidden email]] Im Auftrag von [hidden email]
Gesendet: Dienstag, 20. Juni 2006 18:23
An: [hidden email]
Betreff: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

 


Hai

Thanks so much.  This is what I was looking for.  But I still need some coaching,

[* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.
]

yes, I did this and now I can see classes in the "individual tab". However, it seems I must declare my classes are all subclass of owl:Class in order to be treated as instances.

[Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.
]

This part I didn't get it.

For example, I have class A and B, and I want to say

:A  owl:sameAs :B,

Now I made them subclasses of owl:Class, so I can see them in individual tab.

I also made owl:sameAs property visible.  So I can also see them in as instances of rdf:Property in the individual tab.

What steps I shall follow in order to declare the triple:

:A  owl:sameAs :B


Furthermore, if I want to say a simple triple

:a  :p  :b

where :a is an instance of class A and :b is an instance of class B, :p is an owl:ObjectProperty, without declaring domain and range on property p, I could not find property p in the individual editor.

I am just one step away but could not find the path to go there :)


Thanks  

Helen



Hai Wang <[hidden email]>
Sent by: [hidden email]

06/20/2006 05:21 AM

Please respond to
[hidden email]

To

[hidden email]

cc

 

Subject

[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

 

 

 





Hi Helen,
Protege is an OWL editor, and there is no problem for you to edit OWL-
FULL ontology. You just can not reasoning the OWL-FULL ontology you  
created.
To work on OWL FULL,  you need
* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.

Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.

Regards
Hai

On 6/19/06, [hidden email] <[hidden email]> wrote:
>
> Hai
>
> Thanks for your help.
>
> If for some "strange" reasons we stay with the OWL-FULL ontology,  
am I right
> to conclude that protege owl plug-in is not the tool to author such
> ontology?
>
> My experience is this: although when starting a new project, you  
are asked
> to select the flavour of owl (Lite, DL, Full), even if we select  
OWL FULL,
> the protege editor does not allow one to enter the OWL-FULL type of
> relationship, such as stated in my example:
>
> [
>
>  ProcedureA  hasExpectedOutcome  OutcomeA.
>
>  where ProcedureA and OutcomeA are classes, hasExpectedOutcome  is a
> rdf:Property.
> ]
>
> I was hoping that someone can show me how to enter such OWL-FULL
> relationship using Protege owl plug-in.  Otherwise, I will have to  
go back
> to the text editor, which I definitely would rather not to.
>
> Kind regards.
>
> Helen
>
>
>
>
>
>
>
>
>
>
>
> Hai Wang <[hidden email]>
> Sent by: [hidden email]
>
>
> 06/19/2006 06:24 AM
>
>
> Please respond to
>  [hidden email]
>
>
> To [hidden email]
>
> cc
>
> Subject [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
> express owl:sameAs in Protege
>
>
>
>
>
>
> Hi Helen,
>  Thanks for your clarification.  Now I understood your problem.  If
>  you did not bother OWL Full, maybe you could try to define some meta
>  level classes to  propagate the retractions to the class and its
>  subclasses.
>  For example, we define a class :Meta
>
>
>  :B     rdfs:subClassOf  :A
>  :B rdf:typeof :Meta
>  :A rdf:typeof :Meta
>
>  Then you can put the restrictions to Meta. Meta-class is a common
>  used way to create class templates in Protege-Frames model.
>
>  Regards
>  Hai
>
>  On 15 Jun 2006, at 22:13, [hidden email] wrote:
>
>  >
>  > John and Hai
>  >
>  > I apologize for my non-standard OWL syntax.
>  >
>  > I am used to write triples in N3.  So when I said:
>  >
>  > :A :somePropertyP :C,
>  >
>  > It is a simple triple
>  >
>  > A p C
>  >
>  > where A and C are classes.  OWL does not prohibit using such
>  > triples.  But It make the ontology OWL Full
>  >
>  > Same N3 notations are used in my examples.
>  >
>  > Hai, you are right when you have  A -> p some C, you can make
>  > inference as you stated.  But this is not what I was saying.
>  >
>  >
>  > In my example,
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > here, "include" is just a property, which has Male class and
>  > MaleAnt class as its subject and object respectively.  You can
>  > further consider :include to be a inverse of rdfs:subClassOf  
such that
>  >
>  > :include  owl:inverseOf rdfs:subClassOf
>  >
>  > which means :include is a property defines relationship of
>  > "superClassOf", then you will see no rdfs entailment rules nor owl
>  > entailment rules will let you infer
>  >
>  > :john :include :MaleAnt.
>  >
>  >
>  > One reason of getting into this "messy" OWL-FULL business:   we  
are
>  > trying to model medical guideline and protocols in OWL.  We have
>  > decided all concepts should be modelled as classes, not instances.
>  > In some cases, we must say something like:
>  >
>  > ProcedureA  hasExpectedOutcome  OutcomeA.
>  >
>  > where ProcedureA and OutcomeA are classes, hasExpectedOutcome  
is a
>  > rdf:Property.
>  >
>  >
>  > Helen
>  >
>  >
>  >
>  >
>  >
>  > "John Goodwin" <[hidden email]>
>  > Sent by: [hidden email]
>  > 06/15/2006 03:56 PM
>  > Please respond to
>  > [hidden email]
>  >
>  >
>  > To
>  > <[hidden email]>
>  > cc
>  > Subject
>  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
>  > express owl:sameAs in Protege
>  >
>  >
>  >
>  >
>  >
>  >
>  > Just to clarify when you say
>  >
>  > :A :somePropP :C
>  >
>  > do you mean
>  >
>  > A -> p some C (in Manchester OWL syntax)
>  >
>  > or A p C where A and C are treated as instances and we our now in
>  > OWL full?
>  >
>  > John
>  >
>  > ________________________________
>  >
>  > From: [hidden email] on behalf
> of
>  > [hidden email]
>  > Sent: Thu 15/06/2006 18:50
>  > To: [hidden email]
>  > Cc: [hidden email]
>  > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
>  > editing and express owl:sameAs in Protege
>  >
>  >
>  >
>  > Hai
>  >
>  > One common mistake one can easily make is to think a property
>  > assigned to a class must apply to its subclasses and every of its
>  > members, as follows:
>  >
>  > Given
>  > :A  :somePropertyP :C.  (class A has property P with values from
>  > class C)
>  > :B     rdfs:subClassOf  :A  (class B is a subClass of class A)
>  > :x  rdf:type :A  (a is a member of class A)
>  >
>  > then one can get:
>  >
>  > :B  :somePropertyP :C.  (class B also  has property P with values
>  > from class C) - this is wrong
>  > :x  :somePropertyP :C. (member x has property P with values from
>  > class C) - this is wrong too.
>  >
>  > Jos gave a real life example as following:
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > it would be ridiculous to infer that
>  >
>  > :John :include :MaleAnt.
>  >
>  >
>  > Therefore, even though :A and :B are sub-classes of each other,  
the
>  > properties assigned to each of them can not be passed along.  
Hence:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.}
>  >
>  > an owl reasoner can not entail
>  >
>  > {:A :somePropertyQ :D.
>  > :B :somePropertyP :C}.
>  >
>  > The entailment rules in RDFS are quite simple:
>  >
>  > http://www.w3.org/TR/rdf-mt/#RDFSRules
>  >
>  > It is quite different from OO paradigm.  I got confused many times
>  > - still so from time to time :)
>  >
>  >
>  > In you email, were you saying:
>  >
>  > Given the following triples:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.
>  > :F  :somePropertyP :C
>  > }
>  > now you can entail:
>  >
>  > {:A rd

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: AW: Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

helen.chen

Hi, Raj

Thanks for the tip.   You used "owl:Restriction" in order to make the property appear in the individual tab.

My question, though, is that if I simply just want to state a fact (triple) in protege without imposing the "restriction" clause, for example

:a :p :c

or in OWL:


<rdf:RDF xmlns="http://wopeg.he.agfa.be/rules/mysample#"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#">

    <rdfs:Class rdf:about="http://wopeg.he.agfa.be/rules/mysample#A">
    </rdfs:Class>

    <rdfs:Class rdf:about="http://wopeg.he.agfa.be/rules/mysample#B">
    </rdfs:Class>

    <A rdf:about="http://wopeg.he.agfa.be/rules/mysample#a">
        <p rdf:resource="http://wopeg.he.agfa.be/rules/mysample#b"/>
    </A>

    <B rdf:about="http://wopeg.he.agfa.be/rules/mysample#b">
    </B>

    <rdf:Property rdf:about="http://wopeg.he.agfa.be/rules/mysample#p">
    </rdf:Property>
</rdf:RDF>


Notice that I don't have the restriction here.   Can I express this in protege?  As you have seen, for those build-in owl/rdf properties (sameAs, differentFrom),etc, you can not modify any 'domain' or 'range' on them, so they do not show up on your class tab (A or B) or individual tab (a or b) to be linked to a subject or an object.


Kind regards.

Helen





"Mudunuri, Raj" <[hidden email]>
Sent by: [hidden email]

06/20/2006 02:51 PM

Please respond to
[hidden email]

To
<[hidden email]>
cc
Subject
[protege-owl] AW: Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege






Hi Hai,
 
when I did the first step, in the individual browser it says that there are 32 instances (asserted) of rdfs:Class and 11 instances (asserted) of owl:Class, however in the instance browser I can see only 18 instances (asserted) for rdfs:Class and 3 instances (asserted) for owl:Class… where are the rest?
 
Regarding owl:sameAs, I’m still waiting for some one to tell how to use it… I also made this property visible by checking the box, but I can’t see it as an instance of rdfProperty in the individual tab as Helen cud see…
 
Helen, regarding your last question, you can see what you want in the Forms tab… but I don’t know what we can do with this object property pane in the form editor b’cos I can’t select any thing into it as the buttons inside it are inactive, unlike what we have in Individuals editor…
 
If you want to see what you want in the Individuals editor then you must create individuals for the classes by clicking the purple diamond button with a + sign on the top in the instance browser… Once you have an asserted individual (in the instance browser) for your classes and if the class of this individual has a property then that property will be shown in the individual editor… is this what you want?
 
Have a look at the attachment…
 
Cheers,
Raj
 
 



Von: [hidden email] [mailto:[hidden email]] Im Auftrag von [hidden email]
Gesendet:
Dienstag, 20. Juni 2006 18:23
An:
[hidden email]
Betreff:
[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

 

Hai


Thanks so much.  This is what I was looking for.  But I still need some coaching,


[
* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.

]


yes, I did this and now I can see classes in the "individual tab". However, it seems I must declare my classes are all subclass of owl:Class in order to be treated as instances.


[
Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.
]

This part I didn't get it.


For example, I have class A and B, and I want to say


:A  owl:sameAs :B,


Now I made them subclasses of owl:Class, so I can see them in individual tab.


I also made owl:sameAs property visible.  So I can also see them in as instances of rdf:Property in the individual tab.


What steps I shall follow in order to declare the triple:


:A  owl:sameAs :B



Furthermore, if I want to say a simple triple


:a  :p  :b


where :a is an instance of class A and :b is an instance of class B, :p is an owl:ObjectProperty, without declaring domain and range on property p, I could not find property p in the individual editor.


I am just one step away but could not find the path to go there :)



Thanks  


Helen



Hai Wang <[hidden email]>
Sent by: [hidden email]

06/20/2006 05:21 AM


Please respond to
[hidden email]


To
[hidden email]
cc
 
Subject
[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

 


   






Hi Helen,
Protege is an OWL editor, and there is no problem for you to edit OWL-
FULL ontology. You just can not reasoning the OWL-FULL ontology you  
created.
To work on OWL FULL,  you need
* change the language profile to owl full: OWL->Preference->General -
> choose "OWL FULL" under language profile
*make owl:Class visible: OWL menu --> Preferences ... --> Visibility,  
and then check owl:Class. Once that's done, you'll see rdfs:class in  
the class browser. Under it, you'll see owl:class. In the Individuals  
tab, you will see all the owl class as the instances of owl:class.

Then you can create relations among OWL class instances just like all  
other normal OWL instances. You may need to customize your class form  
(using FORM tab) to enter the relation easily.

Regards
Hai

On 6/19/06, [hidden email] <[hidden email]> wrote:
>
> Hai
>
> Thanks for your help.
>
> If for some "strange" reasons we stay with the OWL-FULL ontology,  
am I right
> to conclude that protege owl plug-in is not the tool to author such
> ontology?
>
> My experience is this: although when starting a new project, you  
are asked
> to select the flavour of owl (Lite, DL, Full), even if we select  
OWL FULL,
> the protege editor does not allow one to enter the OWL-FULL type of
> relationship, such as stated in my example:
>
> [
>
>  ProcedureA  hasExpectedOutcome  OutcomeA.
>
>  where ProcedureA and OutcomeA are classes, hasExpectedOutcome  is a
> rdf:Property.
> ]
>
> I was hoping that someone can show me how to enter such OWL-FULL
> relationship using Protege owl plug-in.  Otherwise, I will have to  
go back
> to the text editor, which I definitely would rather not to.
>
> Kind regards.
>
> Helen
>
>
>
>
>
>
>
>
>
>
>
> Hai Wang <[hidden email]>
> Sent by: [hidden email]
>
>
> 06/19/2006 06:24 AM
>
>
> Please respond to
>  [hidden email]
>
>
> To [hidden email]
>
> cc
>
> Subject [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
> express owl:sameAs in Protege
>
>
>
>
>
>
> Hi Helen,
>  Thanks for your clarification.  Now I understood your problem.  If
>  you did not bother OWL Full, maybe you could try to define some meta
>  level classes to  propagate the retractions to the class and its
>  subclasses.
>  For example, we define a class :Meta
>
>
>  :B     rdfs:subClassOf  :A
>  :B rdf:typeof :Meta
>  :A rdf:typeof :Meta
>
>  Then you can put the restrictions to Meta. Meta-class is a common
>  used way to create class templates in Protege-Frames model.
>
>  Regards
>  Hai
>
>  On 15 Jun 2006, at 22:13, [hidden email] wrote:
>
>  >
>  > John and Hai
>  >
>  > I apologize for my non-standard OWL syntax.
>  >
>  > I am used to write triples in N3.  So when I said:
>  >
>  > :A :somePropertyP :C,
>  >
>  > It is a simple triple
>  >
>  > A p C
>  >
>  > where A and C are classes.  OWL does not prohibit using such
>  > triples.  But It make the ontology OWL Full
>  >
>  > Same N3 notations are used in my examples.
>  >
>  > Hai, you are right when you have  A -> p some C, you can make
>  > inference as you stated.  But this is not what I was saying.
>  >
>  >
>  > In my example,
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > here, "include" is just a property, which has Male class and
>  > MaleAnt class as its subject and object respectively.  You can
>  > further consider :include to be a inverse of rdfs:subClassOf  
such that
>  >
>  > :include  owl:inverseOf rdfs:subClassOf
>  >
>  > which means :include is a property defines relationship of
>  > "superClassOf", then you will see no rdfs entailment rules nor owl
>  > entailment rules will let you infer
>  >
>  > :john :include :MaleAnt.
>  >
>  >
>  > One reason of getting into this "messy" OWL-FULL business:   we  
are
>  > trying to model medical guideline and protocols in OWL.  We have
>  > decided all concepts should be modelled as classes, not instances.
>  > In some cases, we must say something like:
>  >
>  > ProcedureA  hasExpectedOutcome  OutcomeA.
>  >
>  > where ProcedureA and OutcomeA are classes, hasExpectedOutcome  
is a
>  > rdf:Property.
>  >
>  >
>  > Helen
>  >
>  >
>  >
>  >
>  >
>  > "John Goodwin" <[hidden email]>
>  > Sent by: [hidden email]
>  > 06/15/2006 03:56 PM
>  > Please respond to
>  > [hidden email]
>  >
>  >
>  > To
>  > <[hidden email]>
>  > cc
>  > Subject
>  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
>  > express owl:sameAs in Protege
>  >
>  >
>  >
>  >
>  >
>  >
>  > Just to clarify when you say
>  >
>  > :A :somePropP :C
>  >
>  > do you mean
>  >
>  > A -> p some C (in Manchester OWL syntax)
>  >
>  > or A p C where A and C are treated as instances and we our now in
>  > OWL full?
>  >
>  > John
>  >
>  > ________________________________
>  >
>  > From: [hidden email] on behalf
> of
>  > [hidden email]
>  > Sent: Thu 15/06/2006 18:50
>  > To: [hidden email]
>  > Cc: [hidden email]
>  > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
>  > editing and express owl:sameAs in Protege
>  >
>  >
>  >
>  > Hai
>  >
>  > One common mistake one can easily make is to think a property
>  > assigned to a class must apply to its subclasses and every of its
>  > members, as follows:
>  >
>  > Given
>  > :A  :somePropertyP :C.  (class A has property P with values from
>  > class C)
>  > :B     rdfs:subClassOf  :A  (class B is a subClass of class A)
>  > :x  rdf:type :A  (a is a member of class A)
>  >
>  > then one can get:
>  >
>  > :B  :somePropertyP :C.  (class B also  has property P with values
>  > from class C) - this is wrong
>  > :x  :somePropertyP :C. (member x has property P with values from
>  > class C) - this is wrong too.
>  >
>  > Jos gave a real life example as following:
>  >
>  > :Male  :include  :MaleAnt.
>  > :John  rdf:typeOf :Male.
>  >
>  > it would be ridiculous to infer that
>  >
>  > :John :include :MaleAnt.
>  >
>  >
>  > Therefore, even though :A and :B are sub-classes of each other,  
the
>  > properties assigned to each of them can not be passed along.  
Hence:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.}
>  >
>  > an owl reasoner can not entail
>  >
>  > {:A :somePropertyQ :D.
>  > :B :somePropertyP :C}.
>  >
>  > The entailment rules in RDFS are quite simple:
>  >
>  > http://www.w3.org/TR/rdf-mt/#RDFSRules
>  >
>  > It is quite different from OO paradigm.  I got confused many times
>  > - still so from time to time :)
>  >
>  >
>  > In you email, were you saying:
>  >
>  > Given the following triples:
>  >
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.
>  > :F  :somePropertyP :C
>  > }
>  > now you can entail:
>  >
>  > {:A rdfs:subClassOf :F.
>  > :B rdfs:subClassOf :F}.
>  >
>  > I found it hard to understand what rules protege uses to get this
>  > inference.   There is no way that :F should be inferred as
>  > superclass of :A.  They could every be disjoint classes.
>  >
>  >
>  > Helen.
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  > Hai Wang <[hidden email]>
>  > Sent by: [hidden email]
>  >
>  > 06/15/2006 06:55 AM
>  > Please respond to
>  > [hidden email]
>  >
>  >
>  > To
>  > [hidden email]
>  > cc
>  > Subject
>  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing  
and
>  > express owl:sameAs in Protege
>  >
>  >
>  >
>  >
>  >
>  >
>  >
>  > Hi Helen,
>  > I totally agree with what you said about the difference between
>  > owl:equivalentClass and owl:sameAs. However, one thing I am not so
>  > sure is that you said:
>  > *****
>  > {:A  :somePropertyP :C.
>  > :B :somePropertyQ :D.
>  > :A owl:equivalentClass :B.}
>  >
>  > an owl reasoner can not entail
>  >
>  > {:A :somePropertyQ :D.
>  > :B :somePropertyP :C}.
>  > *******
>  >
>  > If I am right, since class A and B have same instances, the  
inference
>  > could be made.
>  > For example, the above model says that
>  > 1. any instances of A will have a value for the property P from  
the
>  > class C
>  > 2. A and B have exactly same set of instances,
>  > Therefore, we can same that any instances of B will have a  
value for
>  > property P from the class C as well, which means that
>  > {:B :somePropertyP :C}.
>  >
>  > If fact, if we created a class "F =  p some C", and run  
classifier.
>  > Both A and B will be classified as the subclasses of F.
>  >
>  >
>  > Regards
>  > Hai
>  >
>  >
>  >
>  >
>  > On 14 Jun 2006, at 20:56, [hidden email] wrote:
>  >
>  > >
>  > > Hi, John
>  > >
>  > > Yes, with this triple:
>  > >
>  > > :A owl:equivalentClass :B
>  > >
>  > > You are only saying the extension of class :A and :B are the  
same,
>  > > i.e. class A and class B have same members.  It is perfectly  
fine
>  > > to say
>  > >
>  > > :A owl:equivalentClass :B.
>  > > :A owl:differentFrom :B.
>  > >
>  > > This is because although :A and :B have the same membership,  
they
>  > > are really two different concepts.
>  > >
>  > > Therefore, given
>  > >
>  > > {:A  :somePropertyP :C.
>  > > :B :somePropertyQ :D.
>  > > :A owl:equivalentClass :B.}
>  > >
>  > > an owl reasoner can not entail
>  > >
>  > > {:A :somePropertyQ :D.
>  > > :B :somePropertyP :C}.
>  > >
>  > > But when you say:
>  > >
>  > > :A owl:sameAs :B.
>  > >
>  > > This means :A and :B are really the same thing (uri).   So if :A
>  > > and :B are classes, all members of class A are also members of
>  > > class B, and vice versa.
>  > >
>  > > Furthermore, given:
>  > >
>  > > {:A  :somePropertyP :C.
>  > > :B :somePropertyQ :D.
>  > > :A owl:sameAs :B.}
>  > >
>  > > an owl reason can entail:
>  > >
>  > > {:A :somePropertyQ :D.
>  > > :B :somePropertyP :C}.
>  > >
>  > > I have just learnt the above difference from Jos De Roo today.
>  > > Hope this helps :)
>  > >
>  > > Kind regards.
>  > >
>  > > Helen
>  > >
>  > >
>  > >
>  > >
>  > > "John Goodwin" <[hidden email]>
>  > > Sent by: [hidden email]
>  > > 06/14/2006 10:49 AM
>  > > Please respond to
>  > > [hidden email]
>  > >
>  > >
>  > > To
>  > > <[hidden email]>
>  > > cc
>  > > Subject
>  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
>  > > express owl:sameAs in Protege
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > > You mean with equivalentTo you can only say the classes have the
>  > > same extension? Yes that's true A equivalentTo B does not  
imply A
>  > > is the same concept as B. However, as far as machine  
reasoning goes
>  > > isn't equivalentTo sufficient? It's perhaps more of user  
interface
>  > > issue to state they are the same concept. Could you add some
>  > > property "sameConceptAs"? So you could say:
>  > >
>  > > A equivalentTo B
>  > > A -> sameConceptAs some B
>  > > B -> sameConceptAs some A
>  > > sameConceptAs is transitive and symmetric
>  > >
>  > > Would that then capture the require machine semantics, and  
also the
>  > > knowledge they are the same concept? Or is that a bit messy?
>  > >
>  > > I've never used Euler or CWM...but aren't the semantics for OWL
>  > > Full hard to define? Might be ok in this case. I know OWL 1.1
>  > > allows this sort of punning. To quote:
>  > >
>  > > In OWL 1.1 a name (like Person) can be used as both an  
individual
>  > > and a class and a property. The computational problems that  
would
>  > > arise if this was treated as in RDF are avoided by ensuring  
that no
>  > > aspect of the use of the name as an individual has any effect on
>  > > the meaning of the name as a class. (This treatment is often  
called
>  > > punning.)
>  > >
>  > > With the above change, general properties can be placed on names
>  > > that are used as classes, and there is additional syntactic  
sugar
>  > > to allow the placement of property values in class and property
>  > > axioms, as in Class(Person partial super(Animal) type
(MyClass) value
>  > > (dc:creator peter)) which is syntactic sugar for Class(Person
>  > > partial) plus Individual(Person super(Animal) type(MyClass)  
value
>  > > (dc:creator peter)). (The above example assumes a definition for
>  > > dc:creator as an individual-valued property.)
>  > >
>  > >
>  > > From
> http://www-db.research.bell-labs.com/user/pfps/owl/
>  > overview.html
>  > >
>  > > I've never really ventured into OWL full so can't really help  
too
>  > > much I'm afraid.
>  > >
>  > > John
>  > >
>  > >
>  > > From: [hidden email]
> [mailto:protege-owl-
>  > > [hidden email]] On Behalf Of [hidden email]
>  > > Sent: 14 June 2006 15:15
>  > > To: [hidden email]
>  > > Cc: [hidden email]; protege-owl-bounce@crg-
>  > > gw.Stanford.EDU
>  > > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
>  > > editing and express owl:sameAs in Protege
>  > >
>  > >
>  > > Raj and John
>  > >
>  > > Thanks for your reply.  My experience is exactly what Raj
>  > > described, word for word.
>  > >
>  > > John, here is the reason I want to use owl:equivalentClass:
>  > >
>  > > When trying to map two knowledge bases, I have a
>  > > concept  :GestationalPatient in OntologyA,  which is defined  
as any
>  > > patient from 0 week to the moment of birth.  In OntologyB, I  
have a
>  > > concept :InUteroPatient, which defined as any patient in  
ulterus.
>  > > When talking about human patients, these two concepts are really
>  > > the same concept, and this is what I want to express with
>  > > owl:equivalentClass.
>  > >
>  > > I can either write
>  > >
>  > > :GestationalPatient owl:sameAs :InUteroPatient
>  > >
>  > > or
>  > >
>  > > :GestationalPatient owl:equivalentClass :InUteroPatient
>  > >
>  > > But with the owl:sameAs relationship, I can only say these two
>  > > classes have the same extension, ie, exactly the same  
members.  I
>  > > really want to say these two concepts are the same concept, not
>  > > just happend to have the same members.
>  > >
>  > > We use semantic web reasoners such as euler and cwm, so being  
OWL-
>  > > FULL is not a problem for me.
>  > >
>  > > I read the protege-owl tutorial and was expecting to see  
those owl
>  > > properties displayed when I checked them all on performance-
>  > > >visibility menu, but didn't see anything.
>  > >
>  > > Kind regards.
>  > >
>  > > Helen
>  > >
>  > >
>  > >
>  > > "John Goodwin" <[hidden email]>
>  > > Sent by: [hidden email]
>  > > 06/14/2006 07:56 AM
>  > > Please respond to
>  > > [hidden email]
>  > >
>  > >
>  > >
>  > > To
>  > > <[hidden email]>
>  > > cc
>  > > Subject
>  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during  
editing and
>  > > express owl:sameAs in Protege
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > >
>  > > Answer below...
>  > >
>  > > I read these sentences, of your previous response, without  
giving a
>  > > break in between... I didn't see that these sentences were  
answers
>  > > for two different questions... so I was expecting to see this
>  > > 'equivalentTo' display in the class hierarchy after going to the
>  > > OWL menu and select preferences ;)
>  > >
>  > > Ahh sorry for the confusion!
>  > >
>  > > What I meant with with OWL Properties is, when you go to the OWL
>  > > menu, then select Preferences, and then select the Visibility  
tab,
>  > > and here when you can see all of the Annotation Properties
>  > > selected, but none of the OWL Properties are checked (atleast  
for
>  > > me) so what is supposed to happen when we check and uncheck  
these
>  > > properties? there is something to do with visibilitiy, so I

>  > > expected that they show a widget (pane, or panel) or  
something like
>  > > that for every kind of property... but the only widget shown is
>  > > for 'owl:disjointWith' in the class editor window... but I

>  > > don't know how different is this 'owl:disjointWith' widget
>  > > from the normal Disjoints widget b'cos both are showing the same
>  > > values...
>  > >
>  > >
>  > > I've not really played with or used these preferences so not  
really
>  > > sure I'm afraid.
>  > >
>  > > John
>  > > .
>  > >


sample.owl (541 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

Hai Wang
In reply to this post by helen.chen
Helen,

On 20 Jun 2006, at 17:23, [hidden email] wrote:

>
> Hai
>
> Thanks so much.  This is what I was looking for.  But I still need  
> some coaching,
>
> [* change the language profile to owl full: OWL->Preference->General -
> > choose "OWL FULL" under language profile
> *make owl:Class visible: OWL menu --> Preferences ... --> Visibility,
> and then check owl:Class. Once that's done, you'll see rdfs:class in
> the class browser. Under it, you'll see owl:class. In the Individuals
> tab, you will see all the owl class as the instances of owl:class.
> ]
>
> yes, I did this and now I can see classes in the "individual tab".  
> However, it seems I must declare my classes are all subclass of  
> owl:Class in order to be treated as instances.

Yes, all the classes you created will be the instances of owl:class  
by default. I guess you do not need to explicitly assert your owl  
class as an instance of owl:Class.

>
> [Then you can create relations among OWL class instances just like all
> other normal OWL instances. You may need to customize your class form
> (using FORM tab) to enter the relation easily.]
>
> This part I didn't get it.
>
> For example, I have class A and B, and I want to say
>
> :A  owl:sameAs :B,
>
> Now I made them subclasses of owl:Class, so I can see them in  
> individual tab.
>

There is a difference. If you make :A and :B as the subclasses of  
owl:Class, you means that :A and :B are some meta class (the  
instances of :A and :B will be some OWL Classes).
If you just create :A and :B as the direct subclass of owl:thing in  
Protege, :A and :B are normal OWL class ( the instances of :A and :B  
will be some normal OWL individuals).  In either cases, you can  
see :A and :B as the individuals of owl:Class in individual tab.

> I also made owl:sameAs property visible.  So I can also see them in  
> as instances of rdf:Property in the individual tab.
>
> What steps I shall follow in order to declare the triple:
> :A  owl:sameAs :B
>
>

You could go to the form tab,
*select owl:Class from the hierarchy.
*double-click on the Form Edit Pane and it brings up a dialog box.
*choose the property you want to associate with owl:class form (e.g.  
owl:sameAs) and select the widget type for the that property (e.g,  
MultiResourceWidget)
*you may need to drag the old owl:class form a little bit, as the  
added widget was hidden at back sometime.
*go back to Individuals tab
*select the owl:class
*select :a as the asserted instances of owl:class
*add the :B (the instances of owl:class) as the property value .

> Furthermore, if I want to say a simple triple
>
> :a  :p  :b
>
> where :a is an instance of class A and :b is an instance of class  
> B, :p is an owl:ObjectProperty, without declaring domain and range  
> on property p, I could not find property p in the individual editor.
>
You can configure the form for the class A by following the above  
steps and associate P to class A's form,

if :a and :b are normal owl instances, you can try OWL-DL Individuals  
tab as well.

Please post to the list if you got any difficulties.

Regards
Hai



> I am just one step away but could not find the path to go there :)
>
>
> Thanks
>
> Helen
>
>
>
>
> Hai Wang <[hidden email]>
> Sent by: [hidden email]
> 06/20/2006 05:21 AM
> Please respond to
> [hidden email]
>
>
> To
> [hidden email]
> cc
> Subject
> [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing and  
> express owl:sameAs in Protege
>
>
>
>
>
>
> Hi Helen,
> Protege is an OWL editor, and there is no problem for you to edit OWL-
> FULL ontology. You just can not reasoning the OWL-FULL ontology you
> created.
> To work on OWL FULL,  you need
> * change the language profile to owl full: OWL->Preference->General -
> > choose "OWL FULL" under language profile
> *make owl:Class visible: OWL menu --> Preferences ... --> Visibility,
> and then check owl:Class. Once that's done, you'll see rdfs:class in
> the class browser. Under it, you'll see owl:class. In the Individuals
> tab, you will see all the owl class as the instances of owl:class.
>
> Then you can create relations among OWL class instances just like all
> other normal OWL instances. You may need to customize your class form
> (using FORM tab) to enter the relation easily.
>
> Regards
> Hai
>
> On 6/19/06, [hidden email] <[hidden email]> wrote:
> >
> > Hai
> >
> > Thanks for your help.
> >
> > If for some "strange" reasons we stay with the OWL-FULL ontology,
> am I right
> > to conclude that protege owl plug-in is not the tool to author such
> > ontology?
> >
> > My experience is this: although when starting a new project, you
> are asked
> > to select the flavour of owl (Lite, DL, Full), even if we select
> OWL FULL,
> > the protege editor does not allow one to enter the OWL-FULL type of
> > relationship, such as stated in my example:
> >
> > [
> >
> >  ProcedureA  hasExpectedOutcome  OutcomeA.
> >
> >  where ProcedureA and OutcomeA are classes, hasExpectedOutcome  is a
> > rdf:Property.
> > ]
> >
> > I was hoping that someone can show me how to enter such OWL-FULL
> > relationship using Protege owl plug-in.  Otherwise, I will have to
> go back
> > to the text editor, which I definitely would rather not to.
> >
> > Kind regards.
> >
> > Helen
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > Hai Wang <[hidden email]>
> > Sent by: [hidden email]
> >
> >
> > 06/19/2006 06:24 AM
> >
> >
> > Please respond to
> >  [hidden email]
> >
> >
> > To [hidden email]
> >
> > cc
> >
> > Subject [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
> editing and
> > express owl:sameAs in Protege
> >
> >
> >
> >
> >
> >
> > Hi Helen,
> >  Thanks for your clarification.  Now I understood your problem.  If
> >  you did not bother OWL Full, maybe you could try to define some  
> meta
> >  level classes to  propagate the retractions to the class and its
> >  subclasses.
> >  For example, we define a class :Meta
> >
> >
> >  :B     rdfs:subClassOf  :A
> >  :B rdf:typeof :Meta
> >  :A rdf:typeof :Meta
> >
> >  Then you can put the restrictions to Meta. Meta-class is a common
> >  used way to create class templates in Protege-Frames model.
> >
> >  Regards
> >  Hai
> >
> >  On 15 Jun 2006, at 22:13, [hidden email] wrote:
> >
> >  >
> >  > John and Hai
> >  >
> >  > I apologize for my non-standard OWL syntax.
> >  >
> >  > I am used to write triples in N3.  So when I said:
> >  >
> >  > :A :somePropertyP :C,
> >  >
> >  > It is a simple triple
> >  >
> >  > A p C
> >  >
> >  > where A and C are classes.  OWL does not prohibit using such
> >  > triples.  But It make the ontology OWL Full
> >  >
> >  > Same N3 notations are used in my examples.
> >  >
> >  > Hai, you are right when you have  A -> p some C, you can make
> >  > inference as you stated.  But this is not what I was saying.
> >  >
> >  >
> >  > In my example,
> >  >
> >  > :Male  :include  :MaleAnt.
> >  > :John  rdf:typeOf :Male.
> >  >
> >  > here, "include" is just a property, which has Male class and
> >  > MaleAnt class as its subject and object respectively.  You can
> >  > further consider :include to be a inverse of rdfs:subClassOf
> such that
> >  >
> >  > :include  owl:inverseOf rdfs:subClassOf
> >  >
> >  > which means :include is a property defines relationship of
> >  > "superClassOf", then you will see no rdfs entailment rules nor  
> owl
> >  > entailment rules will let you infer
> >  >
> >  > :john :include :MaleAnt.
> >  >
> >  >
> >  > One reason of getting into this "messy" OWL-FULL business:   we
> are
> >  > trying to model medical guideline and protocols in OWL.  We have
> >  > decided all concepts should be modelled as classes, not  
> instances.
> >  > In some cases, we must say something like:
> >  >
> >  > ProcedureA  hasExpectedOutcome  OutcomeA.
> >  >
> >  > where ProcedureA and OutcomeA are classes, hasExpectedOutcome
> is a
> >  > rdf:Property.
> >  >
> >  >
> >  > Helen
> >  >
> >  >
> >  >
> >  >
> >  >
> >  > "John Goodwin" <[hidden email]>
> >  > Sent by: [hidden email]
> >  > 06/15/2006 03:56 PM
> >  > Please respond to
> >  > [hidden email]
> >  >
> >  >
> >  > To
> >  > <[hidden email]>
> >  > cc
> >  > Subject
> >  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing
> and
> >  > express owl:sameAs in Protege
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  > Just to clarify when you say
> >  >
> >  > :A :somePropP :C
> >  >
> >  > do you mean
> >  >
> >  > A -> p some C (in Manchester OWL syntax)
> >  >
> >  > or A p C where A and C are treated as instances and we our now in
> >  > OWL full?
> >  >
> >  > John
> >  >
> >  > ________________________________
> >  >
> >  > From: [hidden email] on behalf
> > of
> >  > [hidden email]
> >  > Sent: Thu 15/06/2006 18:50
> >  > To: [hidden email]
> >  > Cc: [hidden email]
> >  > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
> >  > editing and express owl:sameAs in Protege
> >  >
> >  >
> >  >
> >  > Hai
> >  >
> >  > One common mistake one can easily make is to think a property
> >  > assigned to a class must apply to its subclasses and every of its
> >  > members, as follows:
> >  >
> >  > Given
> >  > :A  :somePropertyP :C.  (class A has property P with values from
> >  > class C)
> >  > :B     rdfs:subClassOf  :A  (class B is a subClass of class A)
> >  > :x  rdf:type :A  (a is a member of class A)
> >  >
> >  > then one can get:
> >  >
> >  > :B  :somePropertyP :C.  (class B also  has property P with values
> >  > from class C) - this is wrong
> >  > :x  :somePropertyP :C. (member x has property P with values from
> >  > class C) - this is wrong too.
> >  >
> >  > Jos gave a real life example as following:
> >  >
> >  > :Male  :include  :MaleAnt.
> >  > :John  rdf:typeOf :Male.
> >  >
> >  > it would be ridiculous to infer that
> >  >
> >  > :John :include :MaleAnt.
> >  >
> >  >
> >  > Therefore, even though :A and :B are sub-classes of each other,
> the
> >  > properties assigned to each of them can not be passed along.
> Hence:
> >  >
> >  > {:A  :somePropertyP :C.
> >  > :B :somePropertyQ :D.
> >  > :A owl:equivalentClass :B.}
> >  >
> >  > an owl reasoner can not entail
> >  >
> >  > {:A :somePropertyQ :D.
> >  > :B :somePropertyP :C}.
> >  >
> >  > The entailment rules in RDFS are quite simple:
> >  >
> >  > http://www.w3.org/TR/rdf-mt/#RDFSRules
> >  >
> >  > It is quite different from OO paradigm.  I got confused many  
> times
> >  > - still so from time to time :)
> >  >
> >  >
> >  > In you email, were you saying:
> >  >
> >  > Given the following triples:
> >  >
> >  > {:A  :somePropertyP :C.
> >  > :B :somePropertyQ :D.
> >  > :A owl:equivalentClass :B.
> >  > :F  :somePropertyP :C
> >  > }
> >  > now you can entail:
> >  >
> >  > {:A rdfs:subClassOf :F.
> >  > :B rdfs:subClassOf :F}.
> >  >
> >  > I found it hard to understand what rules protege uses to get this
> >  > inference.   There is no way that :F should be inferred as
> >  > superclass of :A.  They could every be disjoint classes.
> >  >
> >  >
> >  > Helen.
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  > Hai Wang <[hidden email]>
> >  > Sent by: [hidden email]
> >  >
> >  > 06/15/2006 06:55 AM
> >  > Please respond to
> >  > [hidden email]
> >  >
> >  >
> >  > To
> >  > [hidden email]
> >  > cc
> >  > Subject
> >  > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during editing
> and
> >  > express owl:sameAs in Protege
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  >
> >  > Hi Helen,
> >  > I totally agree with what you said about the difference between
> >  > owl:equivalentClass and owl:sameAs. However, one thing I am  
> not so
> >  > sure is that you said:
> >  > *****
> >  > {:A  :somePropertyP :C.
> >  > :B :somePropertyQ :D.
> >  > :A owl:equivalentClass :B.}
> >  >
> >  > an owl reasoner can not entail
> >  >
> >  > {:A :somePropertyQ :D.
> >  > :B :somePropertyP :C}.
> >  > *******
> >  >
> >  > If I am right, since class A and B have same instances, the
> inference
> >  > could be made.
> >  > For example, the above model says that
> >  > 1. any instances of A will have a value for the property P from
> the
> >  > class C
> >  > 2. A and B have exactly same set of instances,
> >  > Therefore, we can same that any instances of B will have a
> value for
> >  > property P from the class C as well, which means that
> >  > {:B :somePropertyP :C}.
> >  >
> >  > If fact, if we created a class "F =  p some C", and run
> classifier.
> >  > Both A and B will be classified as the subclasses of F.
> >  >
> >  >
> >  > Regards
> >  > Hai
> >  >
> >  >
> >  >
> >  >
> >  > On 14 Jun 2006, at 20:56, [hidden email] wrote:
> >  >
> >  > >
> >  > > Hi, John
> >  > >
> >  > > Yes, with this triple:
> >  > >
> >  > > :A owl:equivalentClass :B
> >  > >
> >  > > You are only saying the extension of class :A and :B are the
> same,
> >  > > i.e. class A and class B have same members.  It is perfectly
> fine
> >  > > to say
> >  > >
> >  > > :A owl:equivalentClass :B.
> >  > > :A owl:differentFrom :B.
> >  > >
> >  > > This is because although :A and :B have the same membership,
> they
> >  > > are really two different concepts.
> >  > >
> >  > > Therefore, given
> >  > >
> >  > > {:A  :somePropertyP :C.
> >  > > :B :somePropertyQ :D.
> >  > > :A owl:equivalentClass :B.}
> >  > >
> >  > > an owl reasoner can not entail
> >  > >
> >  > > {:A :somePropertyQ :D.
> >  > > :B :somePropertyP :C}.
> >  > >
> >  > > But when you say:
> >  > >
> >  > > :A owl:sameAs :B.
> >  > >
> >  > > This means :A and :B are really the same thing (uri).   So  
> if :A
> >  > > and :B are classes, all members of class A are also members of
> >  > > class B, and vice versa.
> >  > >
> >  > > Furthermore, given:
> >  > >
> >  > > {:A  :somePropertyP :C.
> >  > > :B :somePropertyQ :D.
> >  > > :A owl:sameAs :B.}
> >  > >
> >  > > an owl reason can entail:
> >  > >
> >  > > {:A :somePropertyQ :D.
> >  > > :B :somePropertyP :C}.
> >  > >
> >  > > I have just learnt the above difference from Jos De Roo today.
> >  > > Hope this helps :)
> >  > >
> >  > > Kind regards.
> >  > >
> >  > > Helen
> >  > >
> >  > >
> >  > >
> >  > >
> >  > > "John Goodwin" <[hidden email]>
> >  > > Sent by: [hidden email]
> >  > > 06/14/2006 10:49 AM
> >  > > Please respond to
> >  > > [hidden email]
> >  > >
> >  > >
> >  > > To
> >  > > <[hidden email]>
> >  > > cc
> >  > > Subject
> >  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
> editing and
> >  > > express owl:sameAs in Protege
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > > You mean with equivalentTo you can only say the classes have  
> the
> >  > > same extension? Yes that's true A equivalentTo B does not
> imply A
> >  > > is the same concept as B. However, as far as machine
> reasoning goes
> >  > > isn't equivalentTo sufficient? It's perhaps more of user
> interface
> >  > > issue to state they are the same concept. Could you add some
> >  > > property "sameConceptAs"? So you could say:
> >  > >
> >  > > A equivalentTo B
> >  > > A -> sameConceptAs some B
> >  > > B -> sameConceptAs some A
> >  > > sameConceptAs is transitive and symmetric
> >  > >
> >  > > Would that then capture the require machine semantics, and
> also the
> >  > > knowledge they are the same concept? Or is that a bit messy?
> >  > >
> >  > > I've never used Euler or CWM...but aren't the semantics for OWL
> >  > > Full hard to define? Might be ok in this case. I know OWL 1.1
> >  > > allows this sort of punning. To quote:
> >  > >
> >  > > In OWL 1.1 a name (like Person) can be used as both an
> individual
> >  > > and a class and a property. The computational problems that
> would
> >  > > arise if this was treated as in RDF are avoided by ensuring
> that no
> >  > > aspect of the use of the name as an individual has any  
> effect on
> >  > > the meaning of the name as a class. (This treatment is often
> called
> >  > > punning.)
> >  > >
> >  > > With the above change, general properties can be placed on  
> names
> >  > > that are used as classes, and there is additional syntactic
> sugar
> >  > > to allow the placement of property values in class and property
> >  > > axioms, as in Class(Person partial super(Animal) type
> (MyClass) value
> >  > > (dc:creator peter)) which is syntactic sugar for Class(Person
> >  > > partial) plus Individual(Person super(Animal) type(MyClass)
> value
> >  > > (dc:creator peter)). (The above example assumes a definition  
> for
> >  > > dc:creator as an individual-valued property.)
> >  > >
> >  > >
> >  > > From
> > http://www-db.research.bell-labs.com/user/pfps/owl/
> >  > overview.html
> >  > >
> >  > > I've never really ventured into OWL full so can't really help
> too
> >  > > much I'm afraid.
> >  > >
> >  > > John
> >  > >
> >  > >
> >  > > From: [hidden email]
> > [mailto:protege-owl-
> >  > > [hidden email]] On Behalf Of [hidden email]
> >  > > Sent: 14 June 2006 15:15
> >  > > To: [hidden email]
> >  > > Cc: [hidden email]; protege-owl-bounce@crg-
> >  > > gw.Stanford.EDU
> >  > > Subject: [protege-owl] Re: AW: Re: AW: Switch to OWL-Full  
> during
> >  > > editing and express owl:sameAs in Protege
> >  > >
> >  > >
> >  > > Raj and John
> >  > >
> >  > > Thanks for your reply.  My experience is exactly what Raj
> >  > > described, word for word.
> >  > >
> >  > > John, here is the reason I want to use owl:equivalentClass:
> >  > >
> >  > > When trying to map two knowledge bases, I have a
> >  > > concept  :GestationalPatient in OntologyA,  which is defined
> as any
> >  > > patient from 0 week to the moment of birth.  In OntologyB, I
> have a
> >  > > concept :InUteroPatient, which defined as any patient in
> ulterus.
> >  > > When talking about human patients, these two concepts are  
> really
> >  > > the same concept, and this is what I want to express with
> >  > > owl:equivalentClass.
> >  > >
> >  > > I can either write
> >  > >
> >  > > :GestationalPatient owl:sameAs :InUteroPatient
> >  > >
> >  > > or
> >  > >
> >  > > :GestationalPatient owl:equivalentClass :InUteroPatient
> >  > >
> >  > > But with the owl:sameAs relationship, I can only say these two
> >  > > classes have the same extension, ie, exactly the same
> members.  I
> >  > > really want to say these two concepts are the same concept, not
> >  > > just happend to have the same members.
> >  > >
> >  > > We use semantic web reasoners such as euler and cwm, so being
> OWL-
> >  > > FULL is not a problem for me.
> >  > >
> >  > > I read the protege-owl tutorial and was expecting to see
> those owl
> >  > > properties displayed when I checked them all on performance-
> >  > > >visibility menu, but didn't see anything.
> >  > >
> >  > > Kind regards.
> >  > >
> >  > > Helen
> >  > >
> >  > >
> >  > >
> >  > > "John Goodwin" <[hidden email]>
> >  > > Sent by: [hidden email]
> >  > > 06/14/2006 07:56 AM
> >  > > Please respond to
> >  > > [hidden email]
> >  > >
> >  > >
> >  > >
> >  > > To
> >  > > <[hidden email]>
> >  > > cc
> >  > > Subject
> >  > > [protege-owl] Re: AW: Re: AW: Switch to OWL-Full during
> editing and
> >  > > express owl:sameAs in Protege
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > >
> >  > > Answer below...
> >  > >
> >  > > I read these sentences, of your previous response, without
> giving a
> >  > > break in between... I didn't see that these sentences were
> answers
> >  > > for two different questions... so I was expecting to see this
> >  > > 'equivalentTo' display in the class hierarchy after going to  
> the
> >  > > OWL menu and select preferences ;)
> >  > >
> >  > > Ahh sorry for the confusion!
> >  > >
> >  > > What I meant with with OWL Properties is, when you go to the  
> OWL
> >  > > menu, then select Preferences, and then select the Visibility
> tab,
> >  > > and here when you can see all of the Annotation Properties
> >  > > selected, but none of the OWL Properties are checked (atleast
> for
> >  > > me) so what is supposed to happen when we check and uncheck
> these
> >  > > properties? there is something to do with visibilitiy, so I
> >  > > expected that they show a widget (pane, or panel) or
> something like
> >  > > that for every kind of property... but the only widget shown is
> >  > > for 'owl:disjointWith' in the class editor window... but I
>  >  > > don't know how different is this 'owl:disjointWith' widget
> >  > > from the normal Disjoints widget b'cos both are showing the  
> same
> >  > > values...
> >  > >
> >  > >
> >  > > I've not really played with or used these preferences so not
> really
> >  > > sure I'm afraid.
> >  > >
> >  > > John
> >  > > .
> >  > >
> >  > >
> >  > > This email is only intended for the person to whom it is
> addressed
> >  > > and may contain confidential information. If you have
> received this
> >  > > email in error, please notify the sender and delete this email
> >  > > which must not be copied, distributed or disclosed to any other
> >  > > person.
> >  > >
> >  > > Unless stated otherwise, the contents of this email are
> personal to
> >  > > the writer and do not represent the official view of Ordnance
> >  > > Survey. Nor can any contract be formed on Ordnance Survey's
> behalf
> >  > > via email. We reserve the right to monitor emails and
> attachments
> >  > > without prior notice.
> >  > >
> >  > > Thank you for your cooperation.
> >  > >
> >  > > Ordnance Survey
> >  > > Romsey Road
> >  > > Southampton SO16 4GU
> >  > > Tel: 023 8079 2000
> >  > > http://www.ordnancesurvey.co.uk
> >  > >
> >  > > .
> >  > >
> >  > >
> >  > > This email is only intended for the person to whom it is
> addressed
> >  > > and may contain confidential information. If you have
> received this
> >  > > email in error, please notify the sender and delete this email
> >  > > which must not be copied, distributed or disclosed to any other
> >  > > person.
> >  > >
> >  > > Unless stated otherwise, the contents of this email are
> personal to
> >  > > the writer and do not represent the official view of Ordnance
> >  > > Survey. Nor can any contract be formed on Ordnance Survey's
> behalf
> >  > > via email. We reserve the right to monitor emails and
> attachments
> >  > > without prior notice.
> >  > >
> >  > > Thank you for your cooperation.
> >  > >
> >  > > Ordnance Survey
> >  > > Romsey Road
> >  > > Southampton SO16 4GU
> >  > > Tel: 023 8079 2000
> >  > > http://www.ordnancesurvey.co.uk
> >  > >
> >  > >
> >  > >
> >  >
> >  >
> >
> ----------------------------------------------------------------------
> >  > ---
> >  > To unsubscribe go to
> > http://protege.stanford.edu/community/
> >  > subscribe.html
> >  >
> >  >
> >  >
> >  >
> >  > .
> >  >
> >  >
> >  > This email is only intended for the person to whom it is  
> addressed
> >  > and may contain confidential information. If you have received
> this
> >  > email in error, please notify the sender and delete this email
> >  > which must not be copied, distributed or disclosed to any other
> >  > person.
> >  >
> >  > Unless stated otherwise, the contents of this email are
> personal to
> >  > the writer and do not represent the official view of Ordnance
> >  > Survey. Nor can any contract be formed on Ordnance Survey's  
> behalf
> >  > via email. We reserve the right to monitor emails and attachments
> >  > without prior notice.
> >  >
> >  > Thank you for your cooperation.
> >  >
> >  > Ordnance Survey
> >  > Romsey Road
> >  > Southampton SO16 4GU
> >  > Tel: 023 8079 2000
> >  > http://www.ordnancesurvey.co.uk
> >  >
> >  >
> >  > <winmail.dat>
> >
> >
> ----------------------------------------------------------------------
> --
> -
> >  To unsubscribe go to
> > http://protege.stanford.edu/community/subscribe.html
> >
> >
> >
>
>
> ----------------------------------------------------------------------
> ---
> To unsubscribe go to http://protege.stanford.edu/community/ 
> subscribe.html
>
>

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

[protege-owl] Re: AW: Re: AW: Re: AW: Switch to OWL-Full during editing and express owl:sameAs in Protege

Hai Wang
In reply to this post by Rajverma
Hello Raj,

On 20 Jun 2006, at 19:51, Mudunuri, Raj wrote:

> Hi Hai,
>
>
>
> when I did the first step, in the individual browser it says that  
> there are 32 instances (asserted) of rdfs:Class and 11 instances  
> (asserted) of owl:Class, however in the instance browser I can see  
> only 18 instances (asserted) for rdfs:Class and 3 instances  
> (asserted) for owl:Class… where are the rest?
There are some other system classes. You can see them by going to  
Project menu ->Configure->Options-> Choose "Display hidden Frames".

Regards
Hai

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html