W3C home > Mailing lists > Public > www-rdf-comments@w3.org > January to March 2002

Re: Using rdf reification to nest statements in N3 like contexts

From: Graham Klyne <GK@NineByNine.org>
Date: Thu, 14 Mar 2002 23:43:02 +0000
Message-Id: <>
To: "Seth Russell" <seth@robustai.net>
Cc: "www-rdf-comments" <www-rdf-comments@w3.org>

With the first level of nesting, I agree that the number of triples for N 
statements increases to 4N.

But with the next level of nesting, in the framework I was describing, each 
of those 4N triples itself becomes 4-fold, for a total, of 16N.  Next level 
of nesting gives 64N.  etc.

Sure, there's a lot of redundant information in there, which is my point 
about implementations optimizing the representation.  What I'm trying to 
say here is that the reification quad gives us a way to formally represent 
this kind of information, but that a pragmatic encoding with 1:1 
correspondence can be more efficient.

E.g., consider:

(1)  s p o .

(2)  _:s1 rdf:type rdf:Statement .
      _:s1 rdf:subject s .
      _:s1 rdf:property p .
      _:s1 rdf:object o .

(3)  _:s21 rdf:type rdf:Statement .
      _:s21 rdf:subject _:s1 .
      _:s21 rdf:property rdf:type .
      _:s21 rdf:object rdf:statement .

      _:s22 rdf:type rdf:Statement .
      _:s22 rdf:subject _:s1 .
      _:s22 rdf:property rdf:subject .
      _:s22 rdf:object s .

      _:s23 rdf:type rdf:Statement .
      _:s23 rdf:subject _:s1 .
      _:s23 rdf:property rdf:property .
      _:s23 rdf:object p .

      _:s24 rdf:type rdf:Statement .
      _:s24 rdf:subject _:s1 .
      _:s24 rdf:property rdf:object .
      _:s24 rdf:object o .

(4)  ...etc...


At 01:13 PM 3/14/02 -0800, Seth Russell wrote:
>From: "Graham Klyne" <GK@ninebynine.org>
> > That said, I rather like this idea.  As it happens, I've made some
> > notes about extending the model theory to handle N3 style contexts
> > (which I think map quite nicely to this suggestion):
> >
> >    http://www.ninebynine.org/RDFNotes/UsingContextsWithRDF.html
>Which is much appreciated :)   But in which you say:
>Using this approach, the number of triples will increase exponentially with
>the depth of context nesting.
>Which I don't think is true.  I think it is basically just 4 (maybe 5) times
>the number of triples. The number of triples for any level of nesting is
>just 4N + (if N>1, N)  where N is the number of statements in whatever level
>of nesting past the top.  Note this even provides that all the statements in
>one context are conjunctive.  See mentograph:
>Seth Russell

Graham Klyne
Received on Thursday, 14 March 2002 18:50:01 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:43:59 UTC