Re: channel classification

["james@xmlTree.com" <james@xxxxxxx.xxxx]

Hi group,

Mark said:
> Think of it this way; users will select the *kinds* of content they want by
> selecting appropriate channels. Where classification may come in handy is in
> categorizing the content within the channel's domain. As such, IMHO it will
> be unworkable to use any third-party scheme, whether it's Yahoo, DMoz, Dewey
> Decimal or anything else; nothing will be specific or relevant.

I have put my money where my mouth is and converted xmlTree to run on the Dewey
classification.  There is a reasonable spread of xml content across the categorisation
structure - some lumpiness does occur (i.e. 605 Technical journals, into which I have put
a whole load of technically-oriented but not computer specific weblogs).  Partly this is
due to my lack of experience with the Dewey System, and this will improve over time.

The other reason why such a system will still work, even if the classification is
incorrect at the finer grains, is that the Dewey system will allow you to filter at higher
as well as lower levels (xmlTree has a checkbox "include all leaves on sub-branches").

One other thing about the Dewey System - it is designed for categorizing subjects.  I want
viewers to also be able to see just weblogs, just xml applications, just forum content,
just static xml documents etc, and this can be seen as a separate axis of categorisation
from the Dewey subject.

Carmen said:
> Well, I need them so that my users don't have to wade through 300+ (and
> at some point within a year, 3000+ ?) unclassified items to find the
> kinds of news that they want.