Closed briesenberg07 closed 2 years ago
dct:spatial only needs to be "spatial characteristics" I believe. There's no rdfs:range, just a suggestion. But there's always coverage (in both namespaces, right?) -- that's a possibility. I think I would choose a "closeMatch" in the "mapping" rather than "None." But it's just mapper's judgment! Whatever the "correct" choice is, I think there's a limitation in cdm. I believe there's only spatial and coverage. Maybe it's not a limitation but a liberation: everything goes. There's like a hundred words stating nothing decisive. If we had a registry, you could state everything with accuracy in the "property file": the cdm-dc mapping is a closeMatch. There is a better match (you could state what it is). Etc.
Thank you for sharing your thoughts @gerontakos !
Right, dct:spatial
is simply defined as "spatial characteristics of the resource".
Hmm... I'm not sure what a spatial characteristic is. Is it a location, as I originally interpreted it? Is it something like "cramped" or "crowded"? Is it something else? All of the above?
The fact that it is a subprop of dct:coverage
is helpful. This property has a range that includes dctclass:Period
, dctclass:Jurisdiction
(I feel confident saying that a building is neither a period or a jurisdiction), and dctclass:Location
. "A spatial region or named place."
OK. I'll call a building a "named place." That's good enough for me, for now!
field settings > DC map = spatial
Although it's of limited usefulness for metadata professionals, it's always worth consulting (you probably already have) https://www.dublincore.org/resources/userguide/creating_metadata/ . Also Hillman's precursor from 200, https://www.dublincore.org/specifications/dublin-core/usageguide/ . At least those documents have a stamp-of-authority.