Re: Proposed simplification of datatype expressions

On March 17, Jeff Heflin writes:
> 
> I'm not sure if Ian's proposal rules out the possibility of saying "more
> than 1," but if so, I can think of a number of examples of these:

The proposal would only provide for minCardinality of 0 or 1 and
maxCardinality of 0, 1 or many, so "more than 1" (i.e., a
minCardinality of 2) would not be expressible.

Note that the proposal would simply make cardinality restrictions on
datatype properties be subject to the same constraints in DL that they
already are in Lite.


> - the class of all people with more than one middle name
> - the class of people with more than one alias
> - the class of products with more than one advertised price

These examples are only semi-convincing. Middle names would generally
be ordered, so some more sophisticated representation than a single
property would probably be needed; an alias is probably a name with
some first/last type of structure rather than just a string; for
multiple advertised prices it seems that a structure including the
source of the advertisement would be better than just a number.

> 
> Specific numbers are a little harder. Maybe saying something like a
> triangle has exactly three values for its angle property? You could
> probably come up with a number of examples like this for CAD or
> geo-spatial database applications.

Maybe, but as I mentioned above, the lack of ordering on properties
makes them of limited value in representing multiple values for
dimensions, coordinates and the like.

Anyway, this clearly isn't a make-or-break issue, so if there is
significant opposition, then I am prepared to drop the suggestion, or
at least move it into the discussion period after last call.

Ian


> 
> Jeff
> 
> Peter Crowther wrote:
> > 
> > > From: pat hayes [mailto:phayes@ai.uwf.edu]
> > >
> > > >"people who have at least three values for property x."
> > > >
> > > >where x is a data valued property. I, too, struggle to think
> > > of an example
> > > >where one would actually want to use such a construct (but am open to
> > > >suggestion)....
> > >
> > > People who own three or more homes, people with at least three
> > > sources of income, people with at least three children, people with
> > > with at least three nationalities, people who have travelled to more
> > > than three foreign nations during the last six months,....
> > 
> > All good examples of cardinality contraints to objects; I have to say
> > that I'm not convinced that any of the following are naturally
> > datatypes:
> > 
> > - homes
> > - sources of income
> > - children
> > - nationalities
> > - travel to a given nation during a date range
> > 
> > Nationality, in my view, comes closest; but why is this better as a
> > datatype than an object?
> > 
> >                 - Peter
> 

Received on Tuesday, 18 March 2003 10:02:30 UTC