Closed briesenberg07 closed 6 years ago
Sorry, I think I understand now.
Physical collections are linked to from source resources, digital collections are linked to from web resources.
This would make differentiating between them in the agents file unnecessary. Unless there is a hypothetical situation where someone is only using the collection file?
Right, we don’t need to differentiate, there is already differentiation. However we have the “files” partly because we want to have the option of adding to them as we see fit. The Collections file would make more sense if it was one continuous collection dataset for all UW Libraries collections. So as we go on – but early in the process – we should probably combine all “collections” files.
So we could add whatever we want to the collection dataset.
What I think we do need to do there is to have only one URI for each entity. Right now some entities, or collections in this case, have multiple URIs: http://doi.org/10.6069/uwlib.55.A.3.4#Alaska-Yukon-PacificExpositionCollection http://doi.org/10.6069/uwlib.55.A.3.4#Alaska-Yukon-Pacific-ExpositionCollection
The hard part about this is that the URIs cleanes in the collection dataset also need to be cleaned in the sourceresource and webresource files.
--Theo
From: Benjamin Riesenberg [mailto:notifications@github.com] Sent: Thursday, July 26, 2018 1:48 PM To: UniverityOfWashingtonLibrariesSchemas/schemasProject schemasProject@noreply.github.com Cc: Subscribed subscribed@noreply.github.com Subject: Re: [UniverityOfWashingtonLibrariesSchemas/schemasProject] Need additional properties in the dcmi:Collection file? (#7)
Sorry, I think I understand now. Physical collections are linked to from source resources, digital collections are linked to from web resources. This would make differentiating between them in the agents file unnecessary. Unless there is a hypothetical situation where someone is only using the collection file?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/UniverityOfWashingtonLibrariesSchemas/schemasProject/issues/7#issuecomment-408230291, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AGokqIVVct6mb9q59JVQzh0SoaCSnaQ3ks5uKisQgaJpZM4ViYuQ.
What I think we do need to do there is to have only one URI for each entity. Right now some entities, or collections in this case, have multiple URIs: http://doi.org/10.6069/uwlib.55.A.3.4#Alaska-Yukon-PacificExpositionCollection http://doi.org/10.6069/uwlib.55.A.3.4#Alaska-Yukon-Pacific-ExpositionCollection
Based on XML metadata, these are actually two entities:
Closing this issue because:
Physical and digital collections appear the same in the collection file. dpla:SourceResources link to dcmi:Collections via dct:isPartOf regardless of whether collections are digital or physical. Additionally, while the AYP digital collection has dct:title "Alaska-Yukon-Pacific Exposition Collection", there is also a physical collection with this same name.
So, should we use additional properties in the Collection file to differentiate between digital and physical collections? The DPLA MAP uses both dct:description and edm:isShownAt to describe dcmi:Collection resources. We could use the former for physical collections and both for digital collections?