Hello!
We run a preprint server that uses OPS as platform – SciELO Preprints (https://preprints.scielo.org/) – and we are currently having a single PDF (galley) per preprint. Therefore we have a single DOI for preprint.
If an author submits a preprint that has a translation (2 or more Galleys/PDF) we ask them to make a new submission for the translation alone, so that the translation would have its own DOI. This is what we’ve been doing so far.
However, we recently found out that OPS, much like OJS, allows DOI assignment at the Galley level, not only at the submission level.
What’s the best way to deal with translation DOI for preprints?
Option one:
Submission - DOI 1
Galley 1 (original language) - DOI 2
Galley 2 (translation) - DOI 3
Option two:
Submission - no DOI
Galley 1 (original language) - DOI 1
Galley 2 (translation) - DOI 2
Option three:
Submission - DOI 1
Galley 1 (original language) - DOI 1 (same as submission DOI)
Galley 2 (translation) - DOI 2
We use Crossref. Does anyone know what would be the best practice here?
Hi there, chiming in a few months later. We use Crossref too and have been in conversation with them about best practices - we’re still actively learning about this. DOIs are of course just document object identifiers. That is, they are a permalink to a certain endpoint. You could apply a DOI to every document you have if you’d like, including your privacy policy and even blog posts, but there is a point of diminishing returns.
My understanding is that Option One would be ideal, or a fourth option where only the submissions have a DOI. The reason would be that to the end user, the submission is the most useful layer of abstraction you may way to link externally to. This means that if you distribute the submission DOI, users will be able to see the full metadata and context before downloading their preferred file. That being said, you can issue DOIs to each of your galleys by all means - the question is just whether that is useful to the user or for internal reasons, or even if it dilutes/confuses citations and reference statistics.
At the end of the day it’s up to you and your organization. Since each org has a DOI prefix, the actual schema and approach you use is up to you entirely. One of the decisions made is what level of abstraction you’d like to append identifiers to.
hello @benaltair !
Thank you for sharing your views.
I realized soon after I posted this, after doing some tests, that Option 1 is currently not possible in OPS given that you can’t assign a DOI to each galley.
In fact, you can, but they are not properly registered in Crossref, so at the end of the day, the DOIs will not work.
I don’t know if that’s a limitation of OPS or Crossref.
Interesting, thanks for sharing some details! We’re using OJS and only apply DOIs for each issue and article (submission). There is that option to add for galleys, but it’s not useful for our needs.
It sounds like that might be a bug with the Crossref Import plugin, as I don’t see any restriction in the Crossref API that would lead to that.