Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.
Sign upComments
This comment has been minimized.
This comment has been minimized.
Checking the first one of these reveals that indeed |
Daniel-Mietchen
changed the title from
Incorrect DOIs stated in PubMed
to
Incorrect DOIs stated in PubMed Central and PubMed
Mar 13, 2018
This comment has been minimized.
This comment has been minimized.
For this case, a Crossref search does not yield any immediate results, so not sure where the problem may come from. Trying another one: In this case, a Crossref search yields a DOI of 10.1016/j.joca.2012.02.324, which points to a chain of redirects currently ending at http://www.oarsijournal.com/article/S1063-4584(12)00396-2/fulltext . |
This comment has been minimized.
This comment has been minimized.
I am focusing right now on removing those erroneous statements from Wikidata. Have done the first batch with the 200 listed above (sample edit). |
This comment has been minimized.
This comment has been minimized.
The next batch is running (sample edit):
|
This comment has been minimized.
This comment has been minimized.
Some of those — e.g. those with DOIs of the form "DOI: 10.4415/ANN_11_04_06" (cf. Q37970965) — look like a simple fix, which I will look into later. I am beginning to think, though, that the best way forward may actually be to keep the wrong DOI statements in Wikidata and to deprecate them there. Given that two reputable sources agree on the values, however, claiming such deprecation would require a reference that attests to the erroneousness of these DOI values, which does not yet exist outside this ticket and the DOI Handbook. |
This comment has been minimized.
This comment has been minimized.
The next batch of 400 (the Wikidata query times out for 500 now, and offsets to get beyond 400 don't work for me either right now):
|
This comment has been minimized.
This comment has been minimized.
Those 400 are now fixed as well (sample edit). |
This comment has been minimized.
This comment has been minimized.
The next 50 (everything larger times out):
|
This comment has been minimized.
This comment has been minimized.
Those 50 are fixed too (sample edit). |
Daniel-Mietchen
changed the title from
Incorrect DOIs stated in PubMed Central and PubMed
to
Incorrect DOIs stated in Europe PMC and PubMed
Mar 13, 2018
This comment has been minimized.
This comment has been minimized.
fnielsen
commented
Mar 13, 2018
Would it be better to deprecate rather than delete? |
This comment has been minimized.
This comment has been minimized.
@fnielsen Yes, deprecation seems to be better, but we need to annotate the deprecation properly (see my comment above), so I think it's OK to remove it for now, but we should eventually put it back in as deprecated, ideally along with whatever valid DOIs these articles actually do have. Meanwhile, I checked the Wikidata query again, and while it timed out in the Wikidata Query Service, it ran through in the Linked Data Fragments endpoint, giving 0 results as of now. Actually, re-running it with a limit 1 still yields a result (Q33810881), which I'm thus using to play around with how to indicate the deprecation. This query gives a list of reasons used in "reason for deprecation" (P2241) statements: SELECT DISTINCT ?reason ?reasonLabel WHERE {
?statement pq:P2241 ?reason.
SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
} Of these, error (Q29485) seemed the most plausible, so I used that on Q33810881. |
This comment has been minimized.
This comment has been minimized.
I went randomly through some more entries in the list and have not found any case yet where the PMC/ PubMed entries were correct. |
This comment has been minimized.
This comment has been minimized.
Summary of today's batch edits, as per https://tools.wmflabs.org/quickstatements/#mode=batches :
|
Daniel-Mietchen
added
the
bug
label
Mar 31, 2018
This comment has been minimized.
This comment has been minimized.
Update: LDF just found 16 new cases, all from items that had been started after the above cleanup. |
Daniel-Mietchen
referenced this issue
Mar 31, 2018
Open
Look into generic options for reporting bugs for sites/ tools that do not have a public bug tracker #764
Daniel-Mietchen
referenced this issue
Sep 3, 2018
Open
Bulk annotate wrong DOIs on PubMed using Hypothesis #936
This comment has been minimized.
This comment has been minimized.
For a related discussion, see https://twitter.com/27point7/status/1070818580789383168 . Current counts
|
Daniel-Mietchen commentedMar 13, 2018
The following Wikidata query lists DOIs in Wikidata that do not start with the obligatory "10":
The first 200 results were