search

plazi / treatmentBank

Repository devoted to house keeping of treatmentBank
0 stars 0 forks source link

updating treamentCitation with new target treatment URL? #71

Open myrmoteras opened 1 year ago

myrmoteras commented 1 year ago

How long does it take - or how can one initiate the linking from a treatmentCitation to the cited treatment that has been uplaoded?

E.g. Erinaceus ecaudatus to Tenrec ecaudatus?

gsautter commented 1 year ago

Provided the names and authorities match up, it shouldn't take longer than 10-15 minutes ... which two treatments is this about?

gsautter commented 1 year ago

Provided the names and authorities match up, it shouldn't take longer than 10-15 minutes

Once you release the IMF, that is, as the write-back service (ELS) needs to be able to acquire the lock to put the links in, of course.

myrmoteras commented 1 year ago

this is the protolog https://treatment.plazi.org/id/54778796-FEAB-FEDB-FEAE-FB8714B04174 I just processed and the MSW should be linked to this one (I did it now manually)

and this are citing the one above https://tb.plazi.org/GgServer/srsStats/stats?outputFields=doc.uuid+doc.articleUuid+bib.title+bib.source+tax.genusEpithet+tax.speciesEpithet&groupingFields=doc.uuid+doc.articleUuid+bib.title+bib.source+tax.genusEpithet+tax.speciesEpithet&FP-tax.genusEpithet=Tenrec&FP-tax.speciesEpithet=ecaudatus&format=HTML

gsautter commented 1 year ago

Small wonder this doesn't link automatically ... the original combination is "Erinaceus ecaudatus, Schreber, 1778", but MSW has the current combination "Tenrec ecaudatus (Schreber, 1777)" (MSW1, MSW2) without any mention of the original combination ... MSW3, where I added the original combination and a respective treatment citation, on the other hand, did link automatically: https://tb.plazi.org/GgServer/html/0ECAFC8797AE11A9A14AB13F5CCA0E2B (check Version History)

myrmoteras commented 1 year ago

we need AI ...

gsautter commented 1 year ago

Even AI cannot reasonably predict the pair-up of a basionym with the current combination based upon a match in species epithet and author and a near-match (1777 vs. 1778) in the year if the genus is completely different ... unless we do a GBIF lookup or the like, that is ... not much to train and statistically model there, catalogs are the only chance (in the absence of proper treatment citations, that is).