Greetings,
The Metadata Working Group is currently
working on a new schema version. As part of this
release, we are considering specific requests and
feedback from clients as well as particular problems
posed by metadata submissions. These can suggest bad
documentation, poor choices in controlled lists, or,
even, gaps in coverage. This last category means that
we are presented with an opportunity to create a new
metadata field, and that needs to be a decision taken
cautiously. This situation has occurred recently when
we had a client submit metadata pertaining to the
collection date of the dataset in a field designed for
publication dates.
It got us thinking about providing
support for the collection dates of data. We'd never
done that before. And, we thought, if we support
temporal information, why not geospatial information?
We took the opportunity to poll some clients about the
efficacy of a suggested approach to handling
geospatial information (a form of georss simple), and
we heard back that the earth scientists mostly liked
it, but that a number of them and others were
concerned that we might be duplicating work better
done by others. One person wondered if this
represented a "slippery slope" for DataCite. Were we
losing our way, trying to add more and more fields
that have nothing to do with citation?
We thought we'd check in with the
membership of DataCite and see how you'd like your
metadata schema to evolve. Should it cleave closely to
the mission of data citation, with a slim set of
additional optional elements? Or, would you like to
see it grow and change, perhaps in new ways to
accommodate the needs of various disciplines? Or
something in the middle, which we invite you to
describe.
Here are some considerations to keep in
mind about this possible decision:
·
There is support for basic geospatial and
temporal information in Dublin Core, so this won’t
present a problem for our corresponding DC application
profile.
·
We already have descriptive metadata in
the optional metadata fields, which are designed to
promote discovery, so any new descriptive fields would
add to that ability.
·
Support for geospatial information will
not address location description needs for many
scientific domains. It works best for the earth
sciences.
·
Scientists have metadata approaches of
their own for detailed descriptions of their datasets.
Should we alter our metadata schema entirely so that
we can take advantage of these domain specific
vocabularies?
As reference, here is a link to the
current version of the schema: http://schema.datacite.org/meta/kernel-2.2/index.html
Thank you for your ideas and feedback.
At this point, because time is drawing
short, please direct any feedback to me.
--Joan