RE: RDF semantics: applications, formalism and education

<- - RDF is not sufficiently expressive, as a pure RDF reader that doesn't
<- understand the extensions for (for example) negation may make incorrect
<- deductions about the information saved by a more expressive
<- system such as a
<- description logic.  This means it's impossible for
<- less-expressive systems
<- to pick crumbs from the more expressive ones --- which was, I
<- believe, one
<- of the ideas behind RDF.

Strangely enough, I think that's the first time I've seen anyone describe
the alleged problems in a human-readable form...

Is RDF hobbled without the logic that some people are saying it needs? Ok,
several people have used formal terms to describe these shortcomings (that I
have serious trouble following, BTW). If the thing(s) that are lacking can
be specified in this way, surely it wouldn't be too much more trouble to
suggest some workarounds and/or alternatives.

Q1. what would be the simplest way of modifying RDF to give it a mechanism
that can be used to capture the meaning of negation etc? (if this is what's
needed)
Q2. could this facility be layered on top without losing too much
interoperability?
Q3. why hasn't 1. or 2. been done already?

I get the picture of a group of mechanics standing around a derailed train,
leaning against a crane saying, 'nah, it'll never run like that...'

Received on Monday, 9 April 2001 08:49:48 UTC