> So *use* that expensive metadata! Shove an effective subject browse in
> somebody's face so they can't miss it. Put a big fat "more like this
> one" button on every record that's returned. These are things that
> aren't that hard (or computationally expensive) to do. Now that we're
> doing a better job of both leaning on our vendors and building our own
> systems, we should make sure we're getting as much leverage out of our
> *existing* tools as we can.
Ok, at least in part we're talking faceted navigation, and that's good.
Illuminate the discovery process. But you're making a dangerous assumption
here: expensive metadata is ipso facto more useful metadata. Just turn a
light on our existing metadata, and our users will follow!
I'd rather have LESS EXPENSIVE metadata, more USER-FRIENDLY metadata, and
MORE OF IT. Yes, I believe in hand-hewing, though computer-assisted creation
is a great, great, great thing. (*Assisted.* *Assisted.*) But we spend far
too much on metadata of questionable use to our users.
In particular, feel free to prove me wrong about our users and "effective
subject browse." Amusing anecdotes about arcane language aside (deglutition,
anyone?), our subject system isn't designed to group items at useful
collection/browsing levels.
If we asked users what information they wanted about resources, it might
radically change our metadata priorities. Or at least it would if we were
listening.
Karen G. Schneider
kgs_at_bluehighways.com
Received on Tue Jun 20 2006 - 10:39:35 EDT