Support ROR IDs for author affiliations

[dhimmel]

I ran into @gbilder at the Frankfurt Book Fair. He suggested using ROR IDs for author affiliations. From https://ror.org/about/:

RROR is the Research Organization Registry, a community-led project to develop an open, sustainable, usable, and unique identifier for every research organization in the world.

Note RROR typo (fix proposed in https://github.com/ror-community/ror-site/pull/23).

Currently, affiliations in metadata.yaml are free text and don't use standardized identifiers. It would be cool if authors could use ROR IDs for their affiliations. It seems like going forward, affiliations should be standardized for the best machine readability.

As an example, the ROR ID for UPENN is https://ror.org/00b30xv10:

Now ROR IDs seem to be at the whole organization level, but manuscripts usually contain affiliations at the department level. Not sure if @gblider has advice here? We could still have authors fill in their department info as free text?

[agitter]

This makes the most sense to me as an optional ID in addition to free text affiliations but not as the primary way to specify affiliations. Specifying one or more departments, centers, programs, etc. associated with a larger entity like a university is important. In addition, not all affiliations will have ROR IDs. One of my affiliations, the Morgridge Institute, does not.

Having more machine readability is great though, and I see the big picture benefits of ROR IDs.

[cthoyt]

What about using a wikipedia Q-identifier for an organization if a ROR is not available? Or, perhaps the Wikidata entry also links a ROR id with the P6782 property and can be used to look it up.

[agitter]

Nice suggestion @cthoyt.

You comment also prompted me to check on the ROR status of the Morgridge Institute. I noted in 2019 it did not have a ROR ID, but it does now.

[cthoyt]

@agitter I communicated with ROR directly when I worked at a startup to get them a ROR identifier - they're pretty open to minting identifiers at different levels of organizations, but understandably can not just go out and curate all of them. They have a form to request curations at https://docs.google.com/forms/d/e/1FAIpQLSdJYaMTCwS7muuTa-B_CnAtCSkKzt19lkirAKG4u7umH9Nosg/viewform (this link is listed on https://ror.org/curation/) so if there's any movement towards this in manubot, then I think it would make sense to write a bit of documentation on how users can add more content to ROR (usually starting in Wikidata) so they can best represent themselves. Now that I think of it, I might write a blog post on how I went through this process!

[cthoyt]

I've done a bit of writing on how new users could get their data into Wikidata, and what happens next. https://cthoyt.com/2021/01/17/organization-organization.html. Please let me know if you have any other thoughts that I could include! Eventually, this might be useful to adapt (and make shorter) for Manubot

[agitter]

Thanks for the really clear and thorough explanation!