Re: subsumption of properties?

Timothy Redmond <[email protected]> Mon, 11 Nov 2013 09:36:23 -0800
Newsgroups gmane.comp.misc.ontology.protege.general
Message-ID <[email protected]>
This is a multi-part message in MIME format.
--===============5402675878300110718==
Content-Type: multipart/alternative;
 boundary="------------030406040903070808020100"

This is a multi-part message in MIME format.
--------------030406040903070808020100
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 11/11/2013 08:52 AM, Michael DeBellis wrote:
> I've been using Protege for a while now and one thing I'm still not 
> clear on is when and why do you subclass a property?  All the other 
> capabilities make perfect sense but I've never come across a system 
> where you make sublcasses of properties or relations before. Can 
> someone give a brief couple of sentences that explains this or point 
> me to the appropriate place in the documentation?


A simple example comes from the pizza ontology wherein hasTopping and 
hasBase are sub-object-properties of hasIngredient.

Informally, p is a sub object property of q if for any pair of 
indiivduals, x and y, if p maps x to y then q maps x to y.  This can be 
formalized by introducing the notion of an interpretation of a language 
which is done for owl 2 here:

                http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/


The formal definition of subObjectProperty is given here (the first 
entry at the top of table 6):

               http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/#Object_Property_Expression_Axioms


Data sub properties are similar.  r is a sub data property of s if for 
any individual, x, and any data value, d, if r maps x to d then s maps x 
to d.


>
> One thing that would help is understanding the formal definition of 
> what it means to say one property is a subclass of another. Looking at 
> properties as relations it seems to me a necessary requirement for 
> property X to be a subclass  of property Y is that the domain and 
> range of X must be subsets of the domain and range of Y. But that 
> doesn't seem like a sufficient definition,

This is a necessary but not a sufficient condition.


-Timothy

> I think I could create examples where that was true but it doesn't 
> seem appropriate to have the two properties be sub-properties.
>
> Michael DeBellis
>
>
> _______________________________________________
> protege-discussion mailing list
> [email protected]
> https://mailman.stanford.edu/mailman/listinfo/protege-discussion
>
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03


--------------030406040903070808020100
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">On 11/11/2013 08:52 AM, Michael
      DeBellis wrote:<br>
    </div>
    <blockquote
      cite="mid:[email protected]"
      type="cite">
      <div style="color:#000; background-color:#fff; font-family:times
        new roman, new york, times, serif;font-size:12pt">
        <div>I've been using Protege for a while now and one thing I'm
          still not clear on is when and why do you subclass a
          property?&nbsp; All the other capabilities make perfect sense but
          I've never come across a system where you make sublcasses of
          properties or relations before. Can someone give a brief
          couple of sentences that explains this or point me to the
          appropriate place in the documentation? <br>
        </div>
      </div>
    </blockquote>
    <br>
    <br>
    A simple example comes from the pizza ontology wherein hasTopping
    and hasBase are sub-object-properties of hasIngredient.<br>
    <br>
    Informally, p is a sub object property of q if for any pair of
    indiivduals, x and y, if p maps x to y then q maps x to y.&nbsp; This can
    be formalized by introducing the notion of an interpretation of a
    language which is done for owl 2 here:<br>
    <br>
    <pre>               <a class="moz-txt-link-freetext" href="http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/">http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/</a>
</pre>
    <br>
    The formal definition of subObjectProperty is given here (the first
    entry at the top of table 6):<br>
    <br>
    <pre>              <a class="moz-txt-link-freetext" href="http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/#Object_Property_Expression_Axioms">http://www.w3.org/TR/2012/REC-owl2-direct-semantics-20121211/#Object_Property_Expression_Axioms</a>
</pre>
    <br>
    Data sub properties are similar.&nbsp; r is a sub data property of s if
    for any individual, x, and any data value, d, if r maps x to d then
    s maps x to d. <br>
    <br>
    <br>
    <blockquote
      cite="mid:[email protected]"
      type="cite">
      <div style="color:#000; background-color:#fff; font-family:times
        new roman, new york, times, serif;font-size:12pt">
        <div style="color: rgb(0, 0, 0); font-size: 16px; font-family:
          times new roman,new york,times,serif; background-color:
          transparent; font-style: normal;"><br>
          <span></span></div>
        <div style="color: rgb(0, 0, 0); font-size: 16px; font-family:
          times new roman,new york,times,serif; background-color:
          transparent; font-style: normal;"><span>One thing that would
            help is understanding the formal definition of what it means
            to say one property is a subclass of another. Looking at
            properties as relations it seems to me a necessary
            requirement for property X to be a subclass&nbsp; of property Y
            is that the domain and range of X must be subsets of the
            domain and range of Y. But that doesn't seem like a
            sufficient definition,</span></div>
      </div>
    </blockquote>
    <br>
    This is a necessary but not a sufficient condition.<br>
    <br>
    <br>
    -Timothy<br>
    <br>
    <blockquote
      cite="mid:[email protected]"
      type="cite">
      <div style="color:#000; background-color:#fff; font-family:times
        new roman, new york, times, serif;font-size:12pt">
        <div style="color: rgb(0, 0, 0); font-size: 16px; font-family:
          times new roman,new york,times,serif; background-color:
          transparent; font-style: normal;"><span> I think I could
            create examples where that was true but it doesn't seem
            appropriate to have the two properties be sub-properties.</span></div>
        <div style="color: rgb(0, 0, 0); font-size: 16px; font-family:
          times new roman,new york,times,serif; background-color:
          transparent; font-style: normal;"><br>
          <span></span></div>
        <div style="color: rgb(0, 0, 0); font-size: 16px; font-family:
          times new roman,new york,times,serif; background-color:
          transparent; font-style: normal;"><span>Michael DeBellis</span></div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
protege-discussion mailing list
<a class="moz-txt-link-abbreviated" href="mailto:[email protected]">[email protected]</a>
<a class="moz-txt-link-freetext" href="https://mailman.stanford.edu/mailman/listinfo/protege-discussion">https://mailman.stanford.edu/mailman/listinfo/protege-discussion</a>

Instructions for unsubscribing: <a class="moz-txt-link-freetext" href="http://protege.stanford.edu/doc/faq.html#01a.03">http://protege.stanford.edu/doc/faq.html#01a.03</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>

--------------030406040903070808020100--

--===============5402675878300110718==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
protege-discussion mailing list
[email protected]
https://mailman.stanford.edu/mailman/listinfo/protege-discussion

Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03

--===============5402675878300110718==--