Search Mailing List Archives
[protege-discussion] An interesting pizza... or max and exactly cardinality not working?
maarten at snowlion.nl
Wed Sep 9 03:17:34 PDT 2009
Thanks for the interesting links. Things make more sense now. This open
world semantics is sometimes confusing. However, my main problem I
realise is that I was considering types, instead of individuals e.g.:
1. Pizza and hasTopping exactly 2 CheeseTopping
I understood as:
2. Pizza and hasTopping exactly 2 /types of/ CheeseTopping
Having pizzas with two (disjoint) cheese toppings in a closed
definition, I couldn't figure out why the reasoner couldn't classify the
pizza. So, it is all about individuals. Makes sense (also).
Of course, in 1. you can never be sure that a pizza doesn't have 3
mozzarella toppings of instance.
Thomas Russ wrote:
> On Sep 7, 2009, at 1:58 AM, maarten wrote:
>> Hi all,
>> Studying the pizza.owl, I didn't get cardinalities max or exactly to
>> work the way I expected.
>> An interesting pizza has the following equivalent class: Pizza and
>> hasTopping min 3 Thing
>> When I change this (or create a similar pizza) with 'max' or
>> 'exactly' instead of 'min', I don't get the expected result (e.g. all
>> pizzas with 3 or less toppings, or exactly 3 toppings).
>> I'm quite new to owl modeling and protege 4, so I'm probably missing
> Open world semantics.
> If you search on the protege-owl mailing list archives, this comes up
> quite frequently.
> For example:
> protege-discussion mailing list
> protege-discussion at lists.stanford.edu
> Instructions for unsubscribing:
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the protege-discussion