Closed oggioniale closed 5 years ago
I'm not sure if the vocabularies defined by Geoscience Australia (GA) are suitable to be used as controlled list of values for the Instrument PIDs that we have in mind.
In particular, you suggest to use Association Type as defined by GA to be used for relationType
. But I can't see how many terms defined there would make sense for an instrument. For instance:
collectiveTitle
: common title for a collection of resources where title identifies elements of a series collectively, combined with information about what volumes are available at the source cite.stereoMate
: part of a set of imagery that when used together, provides three-dimensional imageswasGeneratedBy
: the domain resource was generated by the range resource which must have been a temporal event such as a data processing action or a field surveywasInformedBy
: The domain resource was informed by the range resource which must have been an Entity (i.e. not an Agent with agency or an Activity temporal event). This is a looser association than wasDerivedFrom and does not indicate direct derivation, as wasDerivedFrom doesto name only a few terms. What is this supposed to mean for an instrument?
On the other hand, I fail to find any suitable terms in the GA Association Types for some values currently listed in the schema:
IsDescribedBy
: for linking an article describing the instrument and its measurement methods or other supporting documentation,IsComponentOf
: to indicate that this device is a part of larger instrument (GA has isComposedOf, but the inverse relation seem to be missing),HasMetadata
: to link extended metadata that goes beyond of what fits into our schema.This is a major issue, as I would consider these relations to be important for instrument PIDs.
Concerning the IGSN Codelist that you suggest to use for relatedIdentifierType, I don't quite understand what it is. It looks like a wild mixture of identifier types (ARK
, DOI
, Handle
, IGSN
, ISBN
, …) and relation types (ContactPerson
, Distributor
, Funder
, hasDocument
, …). I can't see how the latter might be suitable values for relatedIdentifierType
. More importantly, PIDINST
is missing in this list.
I think @RKrahl makes valid points here. I also fail to see how the linked lists include necessary and sufficient terms for the schema here. Unless there are any objections from @oggioniale I propose to close this pull request without merging.
Close as discussed in today's meeting.
for the proprieties relatedIdentifierType and relationType I added the URL of vocabularies of terms