W3C home > Mailing lists > Public > www-rdf-interest@w3.org > July 2002

R: questions on assertion

From: Franco Salvetti <franco.salvetti@tiscalinet.it>
Date: Wed, 10 Jul 2002 13:58:55 +0200
Message-ID: <000e01c22809$2e8e7560$6b050a3e@hobbit>
To: <www-rdf-interest@w3.org>, <www-rdf-logic@w3.org>, <phayes@ai.uwf.edu>, <giles.hogben@jrc.it>, <DPawson@rnib.org.uk>
Cc: "Alessandro Provetti" <provetti@gongolo.usr.dsi.unimi.it>

Hi,

1) There are two assertion about the domain of a property, there is a
subproperty of it with an own domain, and an assertion about *something*
using the subproperty:

<rdf:Description rdf:about="http//example.org/ont#p">
  <rdfs:domain rdf:resource="http://example.org/ont#A"/>
  <rdfs:domain rdf:resource="http://example.org/ont#B"/>
</rdf:Description>

<rdf:Description rdf:about="http//example.org/ont#q">
  <rdfs:subPropertyOf rdf:resource="http://example.org/ont#p"/>
  <rdfs:domain rdf:resource="http://example.org/ont#C"/>
</rdf:Description>

<rdf:Description rdf:about="http//example.org/ont#something">
  <ont:q rdf:resource="http://example.org/ont#blabla"/>
</rdf:Description>


/* other useful thinks are supposed already stated */


There are many thinks that we have to consider:

1) What is the domain of ont:p?
2) What is the domain of ont:q?
3) How can I verify that ont:something is in the domain of ont:q?

Please do consider every possible situation:

ont:A and ont:B are *disjoint*
ont:A and ont:B have an not empty *intersection*
ont:C could be a rdfs:subClassof ont:A
ont:C could be a rdfs:subClassof ont:B
ont:C could be a rdfs:subClassof (ont:A U ont:B)
ont:C could have some elements outside (ont:A U ont:B)
[more situation]



I think that is important, in order to reject an assertion, to solve this
problem.

In my personal opinion it is not sufficient know that:

<rdf:Description rdf:about="http://example.org/ont#something>
   <rdf:type rdf:resource="http://example.org/ont#C">
</rdf:Description>

to imply that ont:something is in the valid domain of ont:q because ont:C
could be outside the domain of ont:p and I hope that is impossible that a
subproperty had a domain *larger* than its super-property. I am refering to
the definition of subrelation as a subset of a relation.

In my paper experiments I see problems considering that the domain of ont:p
is the union of ont:A and ont:B
or considering their intersection as a domain of ont:p. This because we do
not have a simply way to say *where* is ont:C just using rdf:subClassOf.

The question is general and related to the inheritnace of the domain for a
subproperty in the case in which there are no other assertion about its
domain and the case in which there are other assertion about it.




sincerely
Franco
Received on Wednesday, 10 July 2002 07:50:57 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:54 GMT