Am 13.11.2003 um 22:50 schrieb Darius:
>
>
> I suspect this means applications will eventually use automated 
> logging of
> user-application or HCI performance for statistical analysis so the 
> application
> can adapt itself to specific users and their habits, thus making the 
> acquiring
> of useful metadata even easier. All this since teams, commitments, and
> requirements are such fluid things that no one wants to keep editing 
> the digital
> representations of such groups of people. You're a member of a group 
> and acquire
> roles and identity by the history of what you do, rather than applying 
> a label
> or being listed in a list.

If you ever have access to a Mac: Play a little with DEVONthink.
(http://www.devon-technologies.com/products/devonthink.php)

This is one of those "free form database/supplementary brain" 
applications.
So you can select text anywhere, press a shortcut, and it's copied into 
the brain.
(or drop a file or  import whole directory-trees... it can work with 
.txt, .doc, .rtf, .pdf
and some more)

One aspect of metadata is classification: We happily sort our stuff in 
directories
to later have any chance to find the document we seach for. Or, if we 
have a
document, we'd like to find related stuff.

DEVONthink tries to do this (after some training) automatically.... and 
this even
seems to work. Very nice app, actually.

       Marcus

--
Marcus Denker marcus at ira.uka.de