W3C home > Mailing lists > Public > www-rdf-interest@w3.org > April 2001

RE: N3 contexts vs RDF reification

From: Lee Jonas <lee.jonas@cakehouse.co.uk>
Date: Wed, 25 Apr 2001 11:12:22 +0100
Message-ID: <51ED29F31E20D411AAFD00105A4CD7A7710E@zingiber.cakehouse.co.uk>
To: "'Seth Russell'" <seth@robustai.net>, RDF-IG <www-rdf-interest@w3.org>
Seth Russell [mailto:seth@robustai.net] wrote:

>From: "Lee Jonas" <lee.jonas@cakehouse.co.uk>
>
>> Furthermore, statements are always translated into quadruples: the
current
>> (p,s,o) triple plus the uri of the containing context - i.e. (c,p,s,o);
>
>There is a competing suggestion to use a statement id in the quadruple
>instead of context uri, like  (id,p,s,o) ... see diagram [1]. The context
>URI would then identify a context description (external to the triples)
>which would collect statements in various contexts.  The technique is also
>easily implemented with a relational database.
>
>This method has three distinct advantages:
>
>1) It models the real world where the context of a proposition is in the
eye
>of the beholder and not a property of the proposition.
>2) It allows for the same proposition (triple) to exist in multiple (even
>disjoint) contexts - also a situation that holds in the real world.
>3) It keeps context in the labeled directed graph data model rather than
>forcing us to adapt a different mathematical structure in mid stream.
>
>I think that all the advantages you ascribe to the (c,p,s,o) technique like
>reification can also be attributed to the (id,p,s,o) technique see [2], but
>with the added benefits above.
>
>This was suggested by Graham Klyne in his recent paper [3] and is the basis
>of my work in progress.
>
>[1] http://robustai.net/mentography/contexts.gif
>[2] http://robustai.net/mentography/reification.gif
>[3] http://lists.w3.org/Archives/Public/www-rdf-interest/2001Mar/0021.html
>
>Seth Russell

I like the concept, but does it mean n+1 tuples? e.g.:

(stmtid1,p,s,o)
(stmtid2,references,ctx1,stmtid)

Also, you have to explicitly identify all statement occurrences
consistently, even across documents - this could prove yet another problem
for generating valid rdf.  

Rather than an 'artificial key' for statements, what if the key were
regarded as (p,s,o)?

then 

(c1,p1,s1,o1)
(c2,p1,s1,o1)

represents two contexts refering to the same statement.  The consumer of
statements is free to 'import' them into their own contexts if they wish:
(c3,p1,s1,o1).

I guess it comes down to the old 'statements vs statement occurrences'
chestnut. [1]

[1] http://www-uk.hpl.hp.com/people/bwm/papers/20001221-paper/  (section
2.5)

Just a thought.

regards

Lee
Received on Wednesday, 25 April 2001 06:12:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:49 GMT