Open jdeck88 opened 6 years ago
It all depends on how people use the database. If they treat expeditions as papers (as I probably would) and projects as larger collaborative groupings (like DIPnet), then I think it makes sense to assign the DOI to the expedition regardless of time. If they treat projects as papers, then it makes sense to assign the DOI to the project regardless of time.
I guess what I am saying is that the unit of data publication is the data underlying a particular paper. I have a feeling that people would be confused if it were otherwise.
Discussion with RJ and Cynthia a few days ago. DOI assignment fits most naturally with a particular project. This is the particular entity that will be most useful to cite and not individual expeditions. Also, expedition loads already have ARK identifiers so the addition of a DOI at the project level makes the most sense. Two ways to approach this: 1) allow users to enter their own DOI, or 2) mint a DOI automatically on project creation, or 3) a combination of the two approaches.
Conclusion:
1) mint a DOI for all new projects and then point this DOI to the project page. When we create the DOI we will point it at a target which is the project page ,e.g. https://geome-db.org/workbench/project-overview?projectId=72
. This will become the project GUID (or perhaps Project Data DOI). Once we activate this, we will disable the ability to add project Data GUID manually in the current project Data GUID field.
2) For the Expedition GUID (https://n2t.net/ark:/21547/DLY2 ) need to redirect to an expedition landing page. For now, this should be the project page. Related to issue #485
While each sample has an ARK we don't offer DOIs for the publication (but each sample has an ARK). While I think its a good idea for Geome to assign DOIs for a data publication we should think about what exactly that is:
I think we're probably talking about assigning DOIs at the project level but wanted to offer up the other alternatives as possibilities also.
Also, while i think a DOI should be issued for a publication that is not mutable (e.g case #4), this would be impractical to implement in our situation. Therefore, i would vote for case #3, which would be easy to implement by assigning a DOI for a project, even if the project's data can change over time.