Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-owl] Problem of "sameAs" working with individuals

Sisi Xuanyuan xuanyuansisi at gmail.com
Tue May 20 13:13:29 PDT 2008


Thanks very much to Russ.

I checked the rule again. There is only one single individual attached to
each individual of "Class_1" and "Class_2". That is because I set the
property "hasProperty" to be functional.

And maybe my naming system has problem. I named one class as "Property", and
just want to say "Class_1" and "Class_2' can associate with "Property"
through the property "hasProperty". So maybe the following is clearer:

Class_1(?c1) and Class_2(?c2) and differentFrom(?c1, ?c2) and Part(?p1) and
Part(?p2) and hasPart(?c1, ?p1)  and hasPart(?c2, ?p2) and sameAs(?p1, ?p2)
--> theExclusiveViolated(?c1, ?c2)

Then I think this rule doesn't seem like a OWL-Full any more.

Looking forward to more advises. Thanks.

Cecilia


------------------------------

Date: Tue, 20 May 2008 08:57:08 -0700
From: Thomas Russ <tar at ISI.EDU>
Subject: Re: [protege-owl] Problem of "sameAs" working with
       individuals
To: User support for the Protege-OWL editor
       <protege-owl at lists.stanford.edu>
Message-ID: <D3AB43CE-4BF7-45DE-B28A-46CD56D97E8F at isi.edu>
Content-Type: text/plain; charset=ISO-2022-JP; format=flowed;
       delsp=yes


>
> Hi all,
>
> I have a problem of "sameAs" when I use it to determining whether
> two individuals are same. The rule is:
>
> Class_1(?c1)  ? Class_2(?c2)  ? differentFrom(?c1, ?c2)
> ? Property(?p1)  ? Property(?p2)  ? hasProperty(?
> c1, ?p1)  ? hasProperty(?c2, ?p2)  ? sameAs(?p1, ?p2)
> ? theExclusiveViolated(?c1, ?c2)
>
> The background is: c1 is an individual of Class_1, c2 is an
> individual of Class_2, and they are different; c1 has property p1,
> c2 has property p2, and both p1 and p2 are individuals of Property;
> I want to check if p1 is same as p2, then the new property
> theExclusiveViolated will be asserted.

This seems to me to be a pretty unusual constraint.  Is it really true
in your domain that each property can only be attached to a SINGLE
individual of the classes 1 and 2?  In that case, it would seem that
you would also run into problems because this rule itself can result
in more than one individual having "theExclusiveViolated" property
asserted about it.

You also realize that by treating properties as individuals (by using
the sameAs property on property individuals, that you have an OWL-Full
ontology.  It may be the case that the SWRL reasoner doesn't handle
classes and properties as individuals.  Perhaps Martin O'Connor can
comment.


> I have already added the list of (p1, p2) and (c1, c2) to an
> allDifferents set. The Jess result shows no property is asserted. If
> I take out the "sameAs(?p1, ?p2)" restriction, this rule works. But
> it is not what I want, so I must add the "sameAs" restriction.



------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/protege-owl/attachments/20080520/aef0565c/attachment.html>


More information about the protege-owl mailing list