I noticed an inconsistent tagging of metadata fields for datasets on CGSpace. We all seem to do it differently!
For example:
cg.identifier.url — the submission form currently says this is not for datasets (!), but for things like blog posts, websites, Google books, etc
cg.identifier.dataurl — the submission form currently hints that this is for "associated" data, but many submitters use this for the link to the dataset (could be a DOI, Handle, or other link)
cg.identifier.doi — some centers add the link here for datasets that have a DOI
After some discussion with the CGSpace team I'm proposing:
Encourage use of cg.identifier.doi as the primary link to datasets if they have one
Encourage use of cg.identifier.url as the primary link to datasets if they do not have a DOI (for example, some Dataverse repositories use Handles and CKAN might use a plain URL)
Encourage use of cg.identifier.dataurl when an item has a related dataset, for example a journal article. This would fit well with metadata fields for other "related" metadata
I noticed an inconsistent tagging of metadata fields for datasets on CGSpace. We all seem to do it differently!
For example:
cg.identifier.url— the submission form currently says this is not for datasets (!), but for things like blog posts, websites, Google books, etccg.identifier.dataurl— the submission form currently hints that this is for "associated" data, but many submitters use this for the link to the dataset (could be a DOI, Handle, or other link)cg.identifier.doi— some centers add the link here for datasets that have a DOIAfter some discussion with the CGSpace team I'm proposing:
cg.identifier.doias the primary link to datasets if they have onecg.identifier.urlas the primary link to datasets if they do not have a DOI (for example, some Dataverse repositories use Handles and CKAN might use a plain URL)cg.identifier.dataurlwhen an item has a related dataset, for example a journal article. This would fit well with metadata fields for other "related" metadata