Re: plural vs singular properties (a proposal)

• From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
• Date: Mon, 23 Apr 2007 09:46:21 -0400 (EDT)
• To: david.price@eurostep.com
• Cc: semantic-web@w3.org
• Message-Id: <20070423.094621.105482769.pfps@research.bell-labs.com>

From: "David Price" <david.price@eurostep.com>
Subject: Re: plural vs singular properties (a proposal)
Date: Mon, 23 Apr 2007 04:13:12 -0500

> Seems like the singular property is closer to the semantics - the relationship 
> is between a Person who is a parent and a Person who is a child, not a Person 
> who is a parent of a SetOfPerson who are children. Contracts that with a 
> Person who is the coach of a SetOfPerson who are a football team.

Yes, one has to be careful about issues like this.

> I'd also suggest that letting how OO programming systems work influence 
> accurate semantics may lead to mistakes. OO programming languages are largely 
> about shared behaviour, not shared semantics (e.g. the "is circle kind of 
> ellipse" debate).

I agree totally with this sentiment.

> Cheers,
> David

It seems to me to be a mistake to have deficiencies in RDFS affect the
modelling process.  

peter

Received on Monday, 23 April 2007 13:46:19 UTC