4 ms·
Doi is the main one. Most published papers and many other things have them. They're newer so often older things hang around too, I'd wager most new pubmed artic
by IanCal 3y ago
Doi is the main one. Most published papers and many other things have them. They're newer so often older things hang around too, I'd wager most new pubmed articles have dois too.
There are several registries, crossref is the big one in the west but it's not the only one. They have probably the best access to the data out of all of the larger registries though.
Dois are pretty good, though not persistent and there's no versioning built in so people have their own formats.
I spent a lot of time working with these as part of https://dimensions.ai https://dimensions.ai for a decade or so. Happy to chat if you want to delve in more.
- atonse 3y agoCool! I know you guys work with OrcID too right? We work with CrossRef to get data but if a DOI is missing, then things get harder to find in CrossRef in our experience.