W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > March 2013

Re: owl:sameAs - Is it used in a right way?

From: Pat Hayes <phayes@ihmc.us>
Date: Tue, 19 Mar 2013 22:11:56 -0500
Cc: Jim McCusker <mccusj@rpi.edu>, Alan Ruttenberg <alanruttenberg@gmail.com>, Erich Gombocz <egombocz@io-informatics.com>, Jeremy J Carroll <jjc@syapse.com>, Umutcan ŞİMŞEK <s.umutcan@gmail.com>, Kingsley Idehen <kidehen@openlinksw.com>, w3c semweb HCLS <public-semweb-lifesci@w3.org>
Message-Id: <CCA99D71-9724-4357-B9A8-4F62670A6BB5@ihmc.us>
To: David Booth <david@dbooth.org>

On Mar 18, 2013, at 5:21 PM, David Booth wrote:

> On 03/18/2013 01:25 AM, Pat Hayes wrote:
>> On Mar 17, 2013, at 8:40 PM, David Booth wrote:
> [ . . . ]
>>> In the semantic web world, these "contextual scopes" are RDF
>>> graphs.
>> No, they aren't. That interpretation of RDF graphs is in direct
>> violation of the RDF specifications. RDF graphs are simply sets of
>> assertions, all in a non-contextual (and non-indexical) logic with no
>> contextual sensitivity.
> I think you're selling your work short.  AFAICT, the notion of context can work beautifully with the existing RDF Semantics.

Contexts can work very well with RDF, but only by changing the semantics, and that change has not been accepted by the RDF WG, so it is not (and will not be) part of the normative standard.

>  One way is to say that the context is the given interpretation, by which an RDF graph's truth-value is to be determined.  That would make that graph's truth-value relative to that interpretation, just as the RDF Semantics currently defines it.

That really does not make sense as a semantic account of contexts. Even with a context to help, assertions such as RDF graphs do not single out *unique* satisfying interpretations. 

>  Another way is to say that the context is the given RDF graph whose truth-value is to be determined.  That would also make the truth-value relative to that RDF graph, just as the RDF Semantics currently defines it.

But it does not define it that way, which is the point. The RDF semantics says that in interpretation is a mapping from names to things in the universe. That mapping is not graph-local: it does not vary from one graph to the next. It is the same for *all* graphs. So it does *not* depend on the context. 

I have worked out how to make RDF into a real context logic. (See http://www.slideshare.net/PatHayes/rdf-with-contexts ) It requires tweaking the truth conditions in the semantics (slide 15). Without this tweak, the contexts don't have any semantic bite. 
Almost exactly the same idea was previously published by Jie Bao, Deb McGuinness & Li DIng, by the way, two years earlier. http://www.slideshare.net/baojie_iowa/2010-0624-rdfcontext.

But, to repeat, this is not currently street-legal RDF. 

> The reason I prefer to think of the context as being the given RDF graph is that (for me) a common usage mode is to start with an RDF graph and use the RDF semantics to determine the set of interpretations that allow that graph to be true.  This allows useful work to be performed.  For example, given an RDF graph containing a URI whose resource identity I wish to determine, I can turn the RDF Semantics crank and find out that that URI must denote an :Apple .

I think that the crank you are turning here is a valid inference crank. BUt that (by itself) doesnt get you contexts. 


> David

IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
Received on Wednesday, 20 March 2013 03:12:31 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:01 UTC