W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > February 2002

reification decision process intro

From: Frank Manola <fmanola@mitre.org>
Date: Thu, 07 Feb 2002 21:14:02 -0500
Message-ID: <3C63346A.7060303@mitre.org>
To: w3c-rdfcore-wg@w3.org
This outlines a suggested plan for organizing the reification discussion 
at our 2002-02-08 telecon, together with some basic input.  Fortunately, 
I could steal all of this from recent email (the question is whether 
I've stolen the right stuff!).  My overall intention is to see if 
everyone (a) agrees on what the decision to be made is and, if so (b) is 
prepared to decide on which option we ought to take.  A possible (c) 
should be some discussion of what to do about the option we *didn't* 
take (e.g., what do we tell the community about it, or how do we modify 
the published documents), but that needs to await (b).

--Frank

Overview:

In http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0064.html 
Brian summarized the issue, saying:

 

 > The formal section of M&S defines rdf:Statement to apply to the 
triple > as Dan pointed out.  However, the examples of reification 
suggest its > use for provenance, for which stating is more useful.
 >
 > I suggest that we don't make a big mountain out of this.  There are
 > two concepts, (subject, predicate, object) and stating.  We have one
 > URI, rdf:Statement.  We pick one concept for rdf:Statement to apply
 > to, and the other will be defined in some other vocabulary(s).  Does
 > it really matter a whole lot which is which?

Process:

In http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0207.html
Jeremy (basically agreeing with Brian on what the issue was) suggested 
the following process, which makes sense to me:


> We now have some idea of the entailments we are discussing.

> I think it is important that we do not vote on these inidividually, and
> potentially end up with a self-contradictory set of statements (logic is not
> democratic).
>
> I think we should instead try to group sets of answer to the entailments in
> two (or maybe three) consistent positions and then have a straight vote to
> decide between them.
>
> e.g.
>
> A Stating reading, for the use-case of provenance, having the following
> entailments hold:
> ???
> and having the following entailments not hold:
> ???
>
> OR
>
> A Statement reading, for maximum consistency with para 162 and 163 of M&S,
> having the following entailments hold:
> ???
> and having the following entailments not hold:
> ???
>
>OR
>
>[...]
>
>
> In summary, I think we should have a statings versus statements show-down,
> using the entailments to help clarify what each position means.

Jeremy also noted in an earlier message that if it turns out that 
provenance and quoting [insert definition here] are NOT supported by 
what we decide reification is, then we need to document some recommended 
solutions for those who want to do those things [I agree;  this is a (c) 
issue].

Test cases/entailments:

I've gleaned the following from the message traffic [add any others you 
like].  Case 1 is the one most of the discussion has been about.  My 
understanding is that a "statement" reading of Case 1 says "yes" to this 
entailment, and that a "stating" [inscription?] reading says "no".

1.  http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0018.html
Does

  <stmt1> <rdf:type>      <rdf:Statement> .
  <stmt1> <rdf:subject>   <subject> .
  <stmt1> <rdf:predicate> <predicate> .
  <stmt1> <rdf:object>    <object> .

  <stmt2> <rdf:type>      <rdf:Statement> .
  <stmt2> <rdf:subject>   <subject> .
  <stmt2> <rdf:predicate> <predicate> .
  <stmt2> <rdf:object>    <object> .

  <stmt1> <property>      <foo> .

  entail:

  <stmt2> <property>      <foo> .

2. http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Feb/0202.html
(Brian) if we decide that:

   <s1> <rdf:type> <rdf:Statment> .
   <s1> <rdf:subject> <subject> .
   <s1> <rdf:predicate> <predicate> .
   <s1> <rdf:object>    <object> .

   <s2> <rdf:type> <rdf:Statment> .
   <s2> <rdf:subject> <subject> .
   <s2> <rdf:predicate> <predicate> .
   <s2> <rdf:object>    <object> .

   <s1> <prop> <value> .

entails

   <s2> <prop> <value> .

then to be consistent we must also decide that anything (and nothing) 
entails:

   _:s <rdf:type> <rdf:Statment> .
   _:s <rdf:subject> <subject> .
   _:s <rdf:predicate> <predicate> .
   _:s <rdf:object>    <object> .

for any subject, predicate and object.

3.     Graham

  <ex:subj> <ex:prop> <ex:obj> .

entails

     _:r <rdf:type> <rdf:Statement> .
     _:r <rdf:subject> <ex:subj> .
     _:r <rdf:predicate> <ex:prop> .
     _:r <rdf:object> <ex:obj> .



-- 
Frank Manola                   The MITRE Corporation
202 Burlington Road, MS A345   Bedford, MA 01730-1420
mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-875
Received on Thursday, 7 February 2002 21:06:01 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:45:08 EDT