W3C home > Mailing lists > Public > semantic-web@w3.org > October 2010

Re: Everything is a Resource

From: Pat Hayes <phayes@ihmc.us>
Date: Thu, 28 Oct 2010 08:05:25 -0500
Cc: semantic-web@w3c.org
Message-Id: <BCBA3216-FC73-4C45-A21D-DC4DBB261707@ihmc.us>
To: Svatopluk Šperka <sperka@gmail.com>

On Oct 25, 2010, at 5:47 PM, Svatopluk Šperka wrote:

> Hi, 
> 
> I have this problem with comprehending how entailment rules are constructed for RDFS concerning "everything is a resource". 
> 
> There are rules rdfs4a and rdf4b which state that subject and object of every triple are resources. 
> 
> These rules seem to me as special cases of rdfs2 and rdfs3 which are saying that subject/object of a triple can be assinged a type according to defined domain/range of a property in a triple. Only condition is that It would have to be given that for every property, its domain and range is class of all resources. 

Yes, but the point of 4a and 4b is to handle the case when this is not said in the RDF graph. That is, the case where the graph gives no explicit information about the domain or range of the property. In this case you still know that the subject and object are resources. 

> 
> Btw. 
> It is noted in RDF Semantics in section 7.3.1 "Extensional Semantic Rules" that they have exactly this consequence (that domain and range of every property is rdfs:Resource).

Not exactly. It is the extensional semantic condition itself which has this consequence. As the text says, those extensional rules are only some of the consequences of this condition. There are others, and these rules are not complete. 

But OK, yes, if we impose the extensional condition onto the RDF semantics, then the resulting semantics makes every triple of the form 

aaa rdfs:range rdfs:Resource .
or
aaa rdfs:domain rdfs:Resource .

always true. Without this condition, they aren't necessarily true (Because even though every individual to which the property applies is in some set, it still doesn't follow that this set is a domain, unless we also have the extensionality condition.)

> Although I don't see how it's possibly from rules that are listed, it's evident from extensional semantic conditions in section 4.2. 
> 
> I would really appreciate if someone could explain rationale behind this to me. 

Rationale behind what, exactly? Im not sure what it is that you are asking about. 

Pat Hayes

> 
> Thank you very much. 
> 
> 
> Svatopluk Šperka 

------------------------------------------------------------
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 Thursday, 28 October 2010 13:07:14 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:41:24 UTC