xml-dev - RE: RDF, the "semantic web", and the nadir of AI (was RE: Realistic pro

[Gavin Thomas Nicol <gtn@ebt.com>]

> The goal as I understang it is "simply" to make the web a
> more usable graph.

That's a lot less than many claim. T

> Right now if you are reading information on the web and want
> access say to related information you have to rely on the
> source to provide with links, or you need to use external
> ad hoc applications (for instance search engines). Both are
> fairly crude ways of relating information, and neither
> really attaches meaning to their targets. RDF, Topic Maps and XLink may
> have different use cases, but they can already be used to better
> what we have.

Of course... though RDF, Topic Maps and XLink all don't *really*
assign meaning to anything though. I should note that between
using <meta> and reverse inference for clustering, clustering
accuracy can be improved tremendously such that "what's related"
as a search isn't too bad. There was a paper on this at one of
the WWW conferences.

> Semantic != complete/correct. Humans can to a certain degree deal with
> incompleteness and inaccuracy. If the meaning is carried clearly enough
> then human interpretation can step in.

The problem is that for meaning to be carried clearly, and for it to
be interpreted clearly, you do need something pretty complete.

> If on the other hand the web was made  more semantic
> (small step by small step) you _may_ eventually be able to look for those
> sources of information based on what you think best defines them ("we rule
> the world, kill all the others, etc").

The problem here is that people need to agree on the classifications,
and they also need to classify things. The classifications themselves
are culturally/socially/temporally dependent.

> But the binding step isn't that hard to make. If you create a language
> within a namespace and bind semantics to its elements, then you could
> describe those semantics (for instance in RDF) and make them available
> either as linked from the document instances, or in some sort of
> repository. It's one (big) step better than relying on the actual name of
> the element, it provides context.

The problem is that there is a "then the magic happens" step here.
There is *no* standard way of defining the semantics in a way that
is globally interpretable, let alone a standard way of binding them
to a namespace. Even if there were it doesn't solve the semantic
evolution problem.

We can do many good things to make the web better, but we should be
careful about overhyping things. The semantic web promises more than
it could ever deliver (I have heard people claim that it will allow
them to ask questions such as "all hotels within a 50 mile radius
of XXX with available rooms at reasonable rates").