- From: Seth Russell <seth@robustai.net>
- Date: Sat, 28 Apr 2001 10:10:52 -0700
- To: "Danny Ayers" <danny@panlanka.net>, <info@jan-winkler.de>, <www-rdf-interest@w3.org>
From: "Danny Ayers" <danny@panlanka.net>
> I don't think if...then... needs to be associated with anything active at
> all - it's only a relationship. Well documented (as a 'production rule')
in
> the expert systems domain. James Tauber's approach with RDF looks good.
The
> inferencing system will make sense of it, presumably along with the rest
of
> the metadata.
But then what is the answer to my question in the example below?
> <- For example: if a RDF model knows {Seth isA Male} and {Male addressedBy
> <- "Mr"} and {[; young Male] addressedBy "boy"}, what is the value of the
> <- <value> ?
Yes, you can certainly represent the mapping of the production rules in RDF.
But, imho, you cannot say "if <class> then <value>" without specifying an
agent and a method ... so that untill you do, that construct would always
remain vague and useless.
Seth
Received on Saturday, 28 April 2001 13:14:52 UTC