Re: [ontolog-forum] Roles in the UBL Trading Cycle

[MDaconta@xxxxxxx]

In a message dated 3/6/2003 1:36:21 AM US Mountain Standard Time, dblack@xxxxxxxxxxxx writes:

> I would tend to model both Buyer and Seller as the names of associations
> that exist between an Order instance and two Parties (I'm introducing a
> commonly used superclass of Organization/Individual).  So an Order (in this
> simplistic example) would have an association to Party called Buyer, and a
> second association to Party called Seller.


Good idea!  You are correct in that treating a role as an Association 
is a good way to capture the fact that this is a behavior and 
transient notion.  I would also say that you attach the buyerOf association
 (or 'buys') to other classes, for example, between Party and Product.

However, when processing a query like:
"Who bought the most apples?"

We need to have the knowledge that to find the answer, we must
know that we search Orders to get to Buyers.  Or, that is where the 
'buyerOf' association comes in handy -- we then search Parties with 
a buyerOf assocation to Apples.

Anyone see problems with this strategy of modeling roles with
associations?

The only thing that I can think of is there could be a problem if you
want to capture additional metadata about a transient notion.  That
would then push it towards modeling it as a class.  So, in this case,
if there are certain properties of Buyer that are exclusive to the act
of buying -- then we need to capture those properties as part of a 
Buyer Class.

- Mike
-------------------------------
Michael C. Daconta
Chief Scientist, Advanced Programs Group
McDonald Bradley, Inc.
www.daconta.net