Important OpenAlex updates

[Jason Priem]

Hi All! Jason from OpenAlex here. We’ve got a bunch of big updates, so I’m bundling them all together in one post: 

• We added about 25M new works! OpenAlex now indexes 237M disambiguated scholarly works. AFAIK, that’s more than any other scientific knowledge graph, save Google Scholar.

• Don’t forget: as noted in the docs in January, the MAG-format snapshot will no longer be supported after July; you’ll need to use the OpenAlex native format snapshot.

• Relatedly: we’re going to start updating the snapshot just once every month, rather than once every fortnight.

• As we improve the quality of MAG’s entity disambiguation, we’re starting to merge entities together. API users won’t notice much change since we’ll serve 302 redirects from redundant IDs to the canonical ones. Snapshot users will need to check the new merged_ids files.

• Last week we (specifically, Jason) presented OpenAlex at the International Conference on the Science of Science & Innovation, and at the NIH. The NIH presentation is on YouTube.

• Our paper was accepted to the STI 2022 conference. Please cite this paper if you use OpenAlex in a research study.

 We’ve got tons of new feature improvements planned for the next month…stay tuned! And as always, please let us know if you find bugs or have ideas for new features!

Best,
Jason

[Mike]

Hi Jason, 

thanks for the info on this major update .

I am writing to seek a confirmation on something I may have misunderstood. 

Is it safe to assume that

(a)  with this release - we now have in OpenAlex "all" records (at least those identified as "works") - from Crossref 

(b) we have all records from PubMed 

reading the paper , this seems to be the case, but can you confirm this ? 

In case you are wondering : this is the context of the work I am doing on assessing OpenAlex data quality . 

Best regards, 

Mike 

[Jason Priem]

My responses are inline below:

On Fri, Jun 24, 2022 at 3:10 AM 'Mike' via OpenAlex users <openale...@googlegroups.com> wrote:

Hi Jason, 

thanks for the info on this major update .

I am writing to seek a confirmation on something I may have misunderstood. 

Is it safe to assume that

(a)  with this release - we now have in OpenAlex "all" records (at least those identified as "works") - from Crossref 

That's *mostly* true. There were a few stragglers that got caught in our update pipeline, and will be added by the end of the month. Also, we purposely have excluded "works" that are actually components, like figures or tables...some journals assign these DOIs, but for nearly all purposes they are not actually works themselves, but parts of other works. We may include them eventually, but only when we've got lots of documentation in place to help folks deal with them.

 

(b) we have all records from PubMed 

reading the paper , this seems to be the case, but can you confirm this ? 

Yes, we should have all records from PubMed.

 

In case you are wondering : this is the context of the work I am doing on assessing OpenAlex data quality . 

Awesome! If you find issues (including missing DOIs or PMIDs), can you please let us know right away? In many cases, we're able to fix those in a week or two, and it helps everyone else who uses the dataset as well. Thanks!

j

 

Best regards, 

Mike 

On Sunday, 19 June 2022 at 00:49:59 UTC+2 wordsl...@gmail.com wrote:

Hi All! Jason from OpenAlex here. We’ve got a bunch of big updates, so I’m bundling them all together in one post: 

• We added about 25M new works! OpenAlex now indexes 237M disambiguated scholarly works. AFAIK, that’s more than any other scientific knowledge graph, save Google Scholar.

• Don’t forget: as noted in the docs in January, the MAG-format snapshot will no longer be supported after July; you’ll need to use the OpenAlex native format snapshot.

• Relatedly: we’re going to start updating the snapshot just once every month, rather than once every fortnight.

• As we improve the quality of MAG’s entity disambiguation, we’re starting to merge entities together. API users won’t notice much change since we’ll serve 302 redirects from redundant IDs to the canonical ones. Snapshot users will need to check the new merged_ids files.

• Last week we (specifically, Jason) presented OpenAlex at the International Conference on the Science of Science & Innovation, and at the NIH. The NIH presentation is on YouTube.

• Our paper was accepted to the STI 2022 conference. Please cite this paper if you use OpenAlex in a research study.

 We’ve got tons of new feature improvements planned for the next month…stay tuned! And as always, please let us know if you find bugs or have ideas for new features!

Best,
Jason

--
You received this message because you are subscribed to the Google Groups "OpenAlex users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openalex-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openalex-users/8d704cb8-5cc8-4094-b4a7-7b209f38370dn%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.