W3C home > Mailing lists > Public > www-webont-wg@w3.org > May 2003

Test Cases: Extended Cardinality Testing

From: Thorsten Liebig <liebig@informatik.uni-ulm.de>
Date: Fri, 16 May 2003 15:47:50 +0200
Message-ID: <3EC4EC06.7020209@informatik.uni-ulm.de>
To: Jeremy Carroll <jjc@hpl.hp.com>
CC: www-webont-wg@w3.org

I found two bugs in section 7.3.4 Extended Cardinality Testing:

The description says:
"This non-entailment can be replicated for any three natural numbers
i, j, k such that i+j < k. In this example, they are chosen as 2, 3 and 6."
but the conclusion is (in Racer syntax):

(define-primitive-concept nonconclusions902
   (and (at-least 2 p)
        (at-least 3 q)
        (at-least 6 r)))

which is indeed satisfiable.
I suppose the following conclusion was intended:

(define-primitive-concept nonconclusions902
   (and (at-least 2 p)
        (at-least 3 q)
        (at-most 4 r)))

But since at-most represents a "less or equal" condition
the description should then say "... i, j, k such that i+j =< k. In this
example, they are chosen as 2, 3 and 4."

The same is to <description-logic/Manifest904#test>.

BTW: the current tests are not very sophisticated. I suggest
to add more complex cardinality test.

E.g. the following entailments:

(define-primitive-concept conclusion-x1
   (and (at-least 2 p)
        (at-most 2 q)
        (exactly 3 r)))

(define-primitive-concept conclusion-x2
   (and (at-most 2 p)
        (at-least 4 r)
        (at-most 2 q)))

(define-primitive-concept conclusion-x3
   (and (at-most 2 p)
        (at-least 4 r)
        (all r A)
        (at-least 3 q)))

Or even more complex:

(define-primitive-concept C)
(define-primitive-concept A C)
(define-primitive-concept B (and C (not A)))
(define-primitive-concept D C)
(define-primitive-role r :range C)
(define-primitive-role p :parent r :range A)
(define-primitive-role q :parent r :range B)
(define-primitive-role s :parent r :range D)

(define-primitive-concept conclusion-x4
   (and (exactly 3 p)
        (exactly 3 q)
        (exactly 3 s)
        (at-most 6 r)))

And some corresponding non-entailments.


Thorsten Liebig              Abteilung Künstliche Intelligenz
Tel.: +49 731 502 4207                Fakultät für Informatik
Fax.: +49 731 502 4119                        Universität Ulm
Received on Friday, 16 May 2003 09:47:55 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:04:45 UTC