Re: [xtm-wg] An Approach to the Semantic Web

[Lars Marius Garshol <larsga@garshol.priv.no>]

* H. Holger Rath
| 
| Two approaches are promissing and could be somehow
| generalized:
| 
| -	Use linguistics software on text documents
| 	and live with the uncertainties and quality 
| 	lacks this approach has.
|
| -	Map XML (SGML) structures to TM constructs
| 	(could be used in conjunction with the
| 	linguistics stuff to find e.g. assocs).
| -	Use intelligent mapping between relational
| 	data in RDBs and TM constructs.

These problems are easy compared to the first one, I would say, since
the structure already exists and it is only a question of mapping the
information into a different data model. It should be possible to come
up with mechanisms that make this relatively painless.

Kal has done much more work on this and a very interesting approach he
calls MDF will be presented at the 'How to Catch a Topic Map' tutorial
in Berlin. Whether it will be him or me presenting it is not yet clear.
(It may also be Steve, I guess.)

If you wish to make use of the textual content of the XML/SGML
documents you are in much the same position as in the first category,
though XML/SGML might be somewhat easier. In any case, natural
language processing is _hard_, but results I've seen from tests of
entity recognition software indicate that useful results can be
obtained. 

--Lars M.


To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/