We had a moderate level (but reliable) identifier delete his account recently. Unfortunately he did not alert any of us prior to doing that, and the region he worked didn’t have a lot of identifiers. So now a few of us are trying to work through there and fix as much as we can.
Does anybody know if there’s a way that I can see what he had done so I can run through those? He had some rare stuff that he ID’d which is now back at higher taxon levels, or wrong ones that his disagreeing ID was keeping it at higher levels. If that’s not possible, has anybody else had this happen? What strategy did you use?
iNat staff have a weekly backup that they can look at, but there’s not much available to the rest of us.
for research grade observations, you might be able to look at the old GBIF DWCA archive files to see who the leading identifier was on a particular observation. for taxa that would be difficult to identify, i would assume that that person’s name should be associated with the identification.
the latest AWS Open Dataset might contain observations, but they won’t say who identified them. so you’d have to basically compare the records there against those in the live system using some other set of criteria – like location and date, or maybe update date (assuming that the account deletion would have caused the observations identified by this user to all get updated at roughly the same time).
Open id tab and filter for the latest updated first, if region doesn’t have a lot going on, observations where ids were deleted from will be the latest updated.
Technically it could be used also for observations of other people but, to avoid disputes, it is limited to the observations of the person running the software.
you could also additionally filter for observations with an older date of upload - thus you might have a higher success rate in finding the affected ones, as those might not have been updated recently, except for the deleted IDs
There are lots of old threads about the process people go through when deleting, and its impacts, I wonder if there’s anything recent that outlines the matter as what I’ve seen doesn’t look very effective. (For example in the screenshots from years back a person gets on deleting tells them lots of numbers about what they’re going to do but doesn’t provide human encouragement for the user to leave the data)… d
