Hi all,
Carl, I enjoyed reading your paper "Accommodating simplicity and complexity in
metadata: lessons from the Dublin Core experience". It certainly identifies some
of the problems we are going to have as communities get down to the
practicalities of implementing metadata schemas - and then trying to share their
metadata! It had always worried me somewhat, the idea of dumbing down things
like author contact details.
Tying this in with the application profiles position paper, I think that
application profiles might be useful for sharing what implementers actually do
(as distinct from the theory and wish lists).
My feeling is that, if the main purpose of your metadata is web resource
discovery, then you should keep as close to simple Dublin Core as you possibly
can. This is what we have done with the HealthInsite gateway. However, many
HealthInsite partners are likely to be contributing to more than one gateway in
the future, each with its own version of Dublin Core / Australian Government
Locator Service metadata schema. Many of these gateways have still not finalised
their schemas. So I am worried that we will continually have to adjust our
harvester software to accommodate the slight differences - dumbing down may not
be enough. For example, one of our partners has just advised that they will be
using 3-character language codes rather than the 2-character ISO ones.
Regards, Prue
Prue Deacon
HealthInsite Editorial Team
Commonwealth Department of Health and Aged Care (Australia)
[log in to unmask]
http://www.healthinsite.gov.au
(HealthInsite metadata specification:
http://www.healthinsite.gov.au/metadata.cfm)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|