W3C home > Mailing lists > Public > public-owl-wg@w3.org > March 2008

Re: nonmon mapping and punning

From: Bijan Parsia <bparsia@cs.man.ac.uk>
Date: Mon, 3 Mar 2008 20:12:44 +0000
Message-Id: <48F830FC-1558-4294-B573-30A63AE5FCCF@cs.man.ac.uk>
To: "Web Ontology Language ((OWL)) Working Group WG" <public-owl-wg@w3.org>

I know there was a later post on this but I couldn't find it :)

On 23 Jan 2008, at 15:20, Jeremy Carroll wrote:

> Peter F. Patel-Schneider wrote:
>
>> For example, making owl11:objectPropertyDomain be a subproperty of
>> rdfs:domain would remove the apparent non-monotonicity in the example
>> given.
>>
> Yes but ...
>
> the concern is what a triple based implementation would actually  
> have to do - and modifying a triple in response to adding a triple  
> would be a fundamental change, that is out-of-kilter with the  
> design principles that we have used to date.

I was studying the Jena documentation today trying to understand this  
point. Prima facie it seemed wrong since Jena is used by many editing  
applications (thus can add and remove triples in response to events  
quite easily) and is supposed to be a general purpose framework (I  
can easily imagine --- and have built! --- systems that discard other  
information upon receiving new information, e.g., a foaf store that  
knew that if I become a friend of your enemy, I'm no longer your  
friend). In any case I found:
	http://jena.sourceforge.net/how-to/event-handler.html
There seem to be some nice examples:
	http://jena.sourceforge.net/javadoc/com/hp/hpl/jena/rdf/listeners/ 
package-summary.html

It seems to me that these facilities could easily be used to handle  
triple at a time updates that had the sorts of effect that the  
current mapping requires. Indeed, wouldn't it work transparently with  
current applications? They wouldn't need to know at all.

In fact, I think some of Jeremy's stronger statements (along the  
lines of "There is no way in Jena to delete a triple when you add  
one", *IIRC*) are wrong...this is exactly such a mechanism.

So, I remain confused on this point. Jeremy, could you clarify?

(I also want to point out that TopBraid Composer is Jena based and  
claims OWL 1.1 support as of the Submission, I believe. Holger, the  
TPC author, talks a lot about using triple oriented toolkits with OWL  
and didn't mention this part of the mapping as a problem, IIRC  
(pointers to places where he did are welcome). I take that as a weak,  
defeasible existence proof that the mapping isn't radically at odds  
with Jena.)

Cheers,
Bijan.
Received on Monday, 3 March 2008 20:10:41 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 3 March 2008 20:10:42 GMT