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

Re: Deciding what to use as the fourth element of the quad

From: Seth Russell <seth@robustai.net>
Date: Fri, 2 Nov 2001 13:02:04 -0800
Message-ID: <00c101c163e1$a3b7aec0$657ba8c0@c1457248a.sttls1.wa.home.com>
To: "RDFIG" <www-rdf-interest@w3.org>
[sorry for the repost, but I used the wrong format for the archives on the
first post]

I am assuming (or rather hoping) that some future version of RDF will
provide us with the mechanism to store and communicate context information.
By "context" I mean a set of triples which can be named and\or described.
For example, the set of triples that I name ContextA is TruthTheoretic,
nonmonotonic and was authored by Jon.  Now this can be implemented 2
different ways that I know of, both of which involves adding an extra
element to our beloved triple.   One way involves calling the fourth element
the URIref of the context, and the other calling the fourth element the
statement id.   I have compared these two different methods in the following
mentograph:

http://robustai.net/mentography/quads2ways.gif

We can also use context to express reification by saying that a reified
statement is a context containing just one statement.  What is interesting
is that the data stored for context-reified statements is the same for both
methods discussed.  I have shown this in the following mentograph:

http://robustai.net/mentography/quads2waysR.gif

Personally I prefer the fourth element to be a context URI ... primarily
because that seems to be the way it is done in CWM.   It also means that we
don't need to grapple with a way to assign URIs to statements, but rather
can assign URIs to contexts ... which me thinks is more useful if we intend
to share them.

What do you all think ?

What is also interesting is that the graphic model, the mentograph, is the
same regardless of what we decide.  This makes me think that mentography is
barking up the right tree.

Seth Russell
http://robustai.net/mentography/
Received on Friday, 2 November 2001 16:07:26 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:52 GMT