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

Re: Sloppy inference rules

From: Pat Hayes <phayes@ihmc.us>
Date: Fri, 23 Nov 2012 16:14:51 -0800
Cc: Tim Berners-Lee <timbl@w3.org>, <nathan@webr3.org>, Richard Cyganiak <richard@cyganiak.de>, Guus Schreiber <guus.schreiber@vu.nl>, RDF WG <public-rdf-wg@w3.org>
Message-Id: <BA22A746-6CBA-4F45-8729-71E4D4DB1BE0@ihmc.us>
To: Jan Wielemaker <J.Wielemaker@vu.nl>

On Nov 22, 2012, at 12:26 AM, Jan Wielemaker wrote:

> I guess we should not (re)start this discussion unless there is
> a concrete proposal to introduce strings as subjects.   My quick
> intuition is that these are a few of the consequences that I don't
> like.
> 
>  - Invites for poor modelling

I have never found the suggested examples compelling, but in any case, this should be handled by good-use guides and tutorials rather than hardwired into the syntax.

>  - Introduces interoperability issues

How so? (Do you mean with legacy systems: well yes, of course. But any change can be objected to on those grounds.)

>  - Harms efficient reasoning.  One of the nice things about
>    resources is that you can compare them quickly.  SPARQL
>    Lit1 = Lit2 is a much more complicated beast.

Incomprehensible. Equality between literals is if anything easier to compute than between IRIs. The problem of establishing owl:sameAs between two IRIs is probably worse then NP-complete. 

Pat

> 
> 	--- Jan
> 
> 
> On 11/21/2012 08:05 PM, Tim Berners-Lee wrote:
>> 
>> Literals in the *subject* position, on the other hand -- are very sensible.
>> I have listed reason for these before, and the discussion must be very
>> old so I'm
>> not inclined to go into them in great depth.
>> 
>> "Fr."  :isShortFor  "France".
>> 
>> 3  expressedAsAString  "3".
>> 
>> 12  :mutuallyPrime 35.
>> 
>> "chat"  :occursIn   :English, :French.
>> Also we have inverses, which make any asymmetry
>> in what can be put in S and O positions lead to strange things,
>> 
>> :foo  :seconds 73.
>> :second owl:inverse :hertz.
>> 
>> for example can be said but leads to an inference which
>> cannot be expressed if you can't put a number as a subject.
>> 
>> Tim
>> 
>> On 2012-11 -19, at 13:20, Jan Wielemaker wrote:
>> 
>>> On 11/19/2012 07:13 PM, Pat Hayes wrote:
>>>> How about literals in subject position in a triple?
>>> 
>>> I think it is the same story.  Invites for
>>> 
>>> "Paris" something:catipalOf "France"
>>> 
>>> While we all know there are other ways to interpret
>>> "Paris".
>> 
>> Yes, but I think your example makes clear:
>> What on earth makes you suppose form this example that
>> there is any difference between the needs for subject and the needs for
>> objects?
>> 
>> Clearly, by this thinking,  literals should due dial allowed in the
>> object position too!
>> 
>> 
>> 
>> 
>> 
>>> I think we are doing right to allow for
>>> literals only in the object position.
>>> 
>>> Cheers --- Jan
>>> 
>> 
> 
> 

------------------------------------------------------------
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 Saturday, 24 November 2012 00:15:24 GMT

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