[topicmapmail] Fwd: CPAN release of WordNet::Similarity

[Murray Altheim]

Bob Parks wrote:
> Greetings:
> Does anyone know of work being done to measure the semantic distance 
> among nodes in a topic map?  Some tools for measuring semantic distance 
> in WordNet were just released to CPAN (below).

Bob,

I'm curious. How does one define "semantic distance", given that
any metric is pretty arbitrary? Is this just a node count between
two words based on whatever existing structure is there in the
thesaurus?

I note that one of the Semantic Web papers I read recently had a
claim that graph distance registered "semantic closeness", yet
when I looked at their example (they used occupations in the UK),
it had nursing right next to marketing. Not very good science
from one of the prime researchers (who will remain unnamed).

[I tend to be very skeptical of any such claims based on my
knowledge of the current state of the art of computational
linguistics, and how quickly people will make inferences that
computers cannot.]

In my own project I have a graph navigator method that can return
a list of nodes within 'n' nodes from a selected one. If the
topic map represents a taxonomy this is an interesting function,
but in places where it's not the function returns very strange
(read: arbitrary) results.

Murray

......................................................................
Murray Altheim                  <http://kmi.open.ac.uk/people/murray/>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK

    Hunt the Boeing! And test your perceptions!
    http://www.asile.org/citoyens/numero13/pentagone/erreurs_en.htm