W3C home > Mailing lists > Public > www-rdf-logic@w3.org > October 2000

Re: semantics of daml:equivalentTo

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Fri, 13 Oct 2000 16:08:30 -0400
To: heflin@cs.umd.edu
Cc: phayes@ai.uwf.edu, www-rdf-logic@w3.org
Message-Id: <20001013160830A.pfps@research.bell-labs.com>
The differences between the various readings of equivalentTo have less to
do with semantics than with with what can be said and what cannot be said.


Suppose you say

SomeRichFriend = (all friend Rich)

[where the above is a macro that expands to the DAML-ONT statments that are
required to make the class SomeRichFriend be precisely those individuals
that have at least one friend that belongs to Rich]

and

<equivalentTo resource="#AllRichFriends" resource="#SomeRichFriend">

are you now (also) allowed to say

<Class about="#AllRichFriends">
  <restrictedBy>
    <Restriction>
      <onProperty resource="#friend"/>
      <toClass resource="#Rich"/>
    </Restriction>
  </restrictedBy>
</Class>

[here I don't need to have a defined class, so I'm giving the straight
DAML-ONT statement]?

If so, then you have a very powerful, but clean, semantic primitive in
equivalentTo.   In this example you would get the ``rule'' that anyone
whose friends are all rich has at least one rich friend, i.e., everyone has
to have a friend.

If not, then you have something that is much less powerful and
it may be possible that ``A <DEF-RENAME> .... can be ignored for the
logical theory''
[http://www.cs.umd.edu/projects/plus/SHOE/pubs/aaai2000.pdf, section 2.2]. 

In both cases, you certainly would not want to have an individual that
belongs to SomeRichFriend but not to AllRichFriends, so it is true that the
``semantic import'' of equivalentTo is that the two classes end up with the 
same meaning.

Peter Patel-Schneider


PS:  A simpler example that shows some of the differences is

	(equivalentTo resource="#Thing" resource="#Nothing")

Does this mean that there can be no individuals?
Received on Friday, 13 October 2000 16:10:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:37 GMT