I believe Andrew has put his finger on one of the great problems for
metadata deployment.
One of the original motivations for the DC workshop series was the
notion that authors could supply their own descriptions, and I certainly
expect the infrastructure to support this option. But I have come to
believe that we are moving from a where-do-I-click mentality to a
who-do-you-trust position. As a representative of the library
community, I see this as an opportunity as much as a problem, given that
the public trust is among our most important assets.
Other formal communities also are positioned to provide trusted resource
description... museums, governmnts, publishers, professional and trade
organizations. There is room for abuse in any such system, and there
will be (already is) in the metadata realm. This just makes it that
much more critical that those with a mission to provide reliable
resource description find common conventions (including means for
validation) on which we may build the future we envision.
stu
-----Original Message-----
From: Andrew Daviel [SMTP:[log in to unmask]]
Sent: Friday, January 23, 1998 3:36 AM
To: meta2
Subject: authentication of metadata
I was at the COMDEX exhibition here (Vancouver) today, and
met very briefly with Paul Flaherty (the originator of
AltaVista) folowing
his keynote address.
I posed the general question "what about metadata?" He said that
they'd
used metadata more initially ("keywords" presumeably) but cut
back
because of word spamming (for the uninitated, this is the
practice of
putting "keywords = microsoft sex drugs music business ...." on
your
pizza page, just to get hits).
He suggested that one might use some kind of crypto
authentication to
validate metadata, but there wasn't time to discuss it properly.
I presume that when his search engine produces irrelevant hits,
he gets
the blame, rather than the author of the page who has written
all this
junk in black-on-black or fontsize -6 or whatever.
Andrew Daviel
|