W3C home > Mailing lists > Public > public-rdf-wg@w3.org > November 2012

Re: Sloppy inference rules

From: Pat Hayes <phayes@ihmc.us>
Date: Mon, 19 Nov 2012 10:13:56 -0800
Cc: nathan@webr3.org, Richard Cyganiak <richard@cyganiak.de>, Guus Schreiber <guus.schreiber@vu.nl>, RDF WG <public-rdf-wg@w3.org>
Message-Id: <89852539-6037-4C45-8B17-72D659FAC43A@ihmc.us>
To: Tim Berners-Lee <timbl@w3.org>

On Nov 19, 2012, at 7:28 AM, Tim Berners-Lee wrote:

> In general,  I think this is going in a better direction.
> 
> 
> 1)  A bnode as a predicate is, I find, sometimes valuable.

I agree.

>  It doesn't change the RDFs of things, it does allow to state things about the predicate without having to invent URIs on the fly which is sometimes (a) a pain to go in running code and (b) a deliberate wish not to make something you have to then support and nurture as a public Property.
> 
> 
> <JoesHomePage.html>   [  sandro:parallelProperty fb:likes  ]  <RockyHorrorShow.html> .
> 
> :Alice  [ is owl:inverseOf  ex:supervisor ]  :Bob.
> 
> 
> Bnode Properties, though not in the predicate slot in the statement, the tabulator
> will use if it find something like:
> 
> :child a property, rdfs:label "parent", owl:inverseOf [rdfs:label "parent" ].
> 
> 
> 
> 2) Literals as predicates? No thanks.

I also agree, this is a minority taste. It is legal is ISO Common Logic, and we did find an actual use for it, but we could just as easily done the job without needing this peculiarity. And it would raise a host of nasty semantics to deal with things like ill-typed literals used as predicates. So, nix to that. 

How about literals in subject position in a triple? 

Pat

> 
> In a view of the world I find convenient, literals and properties
> are disjoint.
> 
> Also I assume
> 
> {   ?s ?p ?o } => { ?p a rdf:Property }.
> 
> So to say
> 
> 	Alice 3 Bob
> 
> would give an inconstancy.  You could say it wasn't the RDF graph level.
> 
> So I'd be content to see parsers which accepted a number as a property I suppose,
> but my system would flag it as an error at the next stage.
> 
> It is good to be able to flag it as an error, as Turtle is getting 
> toward the undesirable state for a language where 
> almost anything you write will parse, and so ability to trap typos is reduced.

Good point. 

> 
> Another reason I would really want to keep predicates distinct form any literals 
> would be in the UI in user input mode I want to give the user, when choosing say a form field or predicate value for a new relationship, a choice of Properties 
> but I really don't want to have to give choices of literal values!
> 
> 
> Tim
> 
> 
> 
> On 2012-11 -07, at 06:23, Nathan wrote:
> 
>> Richard Cyganiak wrote:
>>> On 30 Oct 2012, at 16:15, Pat Hayes wrote:
>>>> So my question to the WG is, how disastrous would it be if the RDFS inference rules were stated in terms of a 'sloppy' version of RDF with the syntactic restrictions removed, i.e. with the following grammar for triples?
>>>> 
>>>> triple ::= term term term .
>>>> term ::=  IRI | blanknode | literal
>>> +1
>> 
>> +1
>> 
>> 
> 
> 

------------------------------------------------------------
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 Monday, 19 November 2012 18:14:32 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:53 GMT