Jama Poulsen wrote:
I agree with this, but the solution isn't to not
use
tags/labels/categories.
The real solution lies in a combination of various
meta-info systems and a
powerful browse/search interface. For a project like Wikipedia a very
flexible
solution is needed.
Semantic Mediawiki is very cool - I'll have to look at that some more.
Another approach relevant to what you're describing is to use multiple tags
for articles, and then search against those multiple categories (described
at
http://meta.wikimedia.org/wiki/Category_math and other places) - one
example of this in action is at
http://www.wikidirectory.org/index.php?title=Special:Intersections&catl…rces.
This gives you the more specific results from a combination of tags or
categories having less specific meaning.
Using an approach where you embrace tags/categories/labels/whatever that are
less specific, you begin to get an "average" kind of meaning. I've also
been thinking about how to combine this approach (category intersections)
with some kind of relevance rating (maybe using the article validation code
to rate the relevance of each tag).
Best Regards,
Aerik