Deleted identifications

Ahhh,yes, you are right. My mistake in getting the two confused. Thanks.

2 Likes

That seems to be the appropriate use of the functionality.
On the taxon page, we have also a section of “similar species”, i.e.“commonly misidentified” species. How does deleting vs. withdrawing an id influence the data shown there?

Deleted IDs are removed from the database, so the data isn’t available for anything anymore. We will use withdrawn and non-current/inactive/older IDs when generating “similar species”.

This is how I view it as well. I’d only delete an identification if it was added unintentionally

5 Likes

Is that true? This observation has one withdrawn ID of Grapsus grapsus and three active IDs of Leptograpsus variegatus.

Grapsus grapsus is not listed under the similar species of Leptograpsus variegatus, and Leptograpsus variegatus is not listed under the similar species of Grapsus grapsus.

2 Likes

It looks like you’re right. I was working off memory, and after reviewing the code it looks like we only use active identifications for the similar species results. I can’t recall if we did at one point use inactive IDs.

My stance on deleted identifications remains - in my view identifications should only be deleted if added unintentionally. Deleted IDs are removed from the system but withdrawn IDs are still available via the API for anyone wanting to use that data.

4 Likes

I’d always assumed withdrawn IDs fed the similar species, but that there had to be a certain threshold met, i.e. a one off wouldn’t be enough to add it, but if, say, 10% of a species total observations were at some point (be initial or later IDs), getting the IDed as another species, even if it was later withdrawn, then it would contribute to similar species.

That seems like an oversight. A withdrawn ID (as opposed to deleted, which this thread seems to informally indicate is predominately for misclicks & typos) means at one point it was confused for the species at some point, right?

I suppose one could argue that active IDs means that someone believes in the ID enough to stand by it and not withdraw, and therefore it is a mistake more likely to be repeated by others…except that doesn’t track, because there are lots of active IDs that belong to inactive accounts, from users who have a last active date in their profile from YEARS past.

If their ID that they left and never returned to check (cough Geralds of the world cough) feeds the similar species, why shouldn’t the ID of someone who was corrected and withdrew count as well?

Also, just curious: does the similar species contribute to the CV (eg, along with what it suggests, it considers the similar species for that guess)?

3 Likes

There actually is a threshold – it’s more than one. So the example I gave is sort of misleading because I only showed one. But there is another observation with these same species, so it would meet the two or more requirement if the withdrawn ID were counted.

1 Like

No. The “similar species” shown on the taxon page aren’t used for anything other than displaying in that section of the taxon page on the website, and a section of taxon pages in the Seek app.

2 Likes

I can see the point of a complete deletion in order to hide the location of an endangered species. As a Plant Conservation Volunteer, I learned to ensure that iNaturalist never gets to share the sight; photos automatically give exact GPS coordinates, and survey sites have to be protected from being poached or trampled. I imagine “withdrawn” would still have those background data, so it’s better to erase the record entirely.

You can obscure or even completely hide GPS coordinates on iNat observations. The metadata is scrubbed on the actual photo, and the website keeps it hidden – either automatically for known endangered species (those that have been flagged as such), or manually if you choose to obscure the location.
If someone else’s observation of an endangered species is not obscured, you should flag the species on iNat and a Curator can turn on auto-obscuration of the specie’s locations.

2 Likes

I think you may be confusing deleting an identification with deleting an entire observation. An identification (the subject of this thread) is just a taxon “label” that any iNat user can attach to an observation to indicate what they think it is. Identifications do not carry any location information – that information is attached directly to the observation.

And as @trh_blue pointed out, an observer can always obscure or hide the location of any observation, using the “obscure” or “private” Geoprivacy setting on their observation. The owner of the observation will still see the exact location, but no one else will (unless you specifically give them permission), and it won’t be available to others from the photos either. (If you want to test that, access the observation URL from a separate web browser where you are not logged in to iNaturalist.)

So there should never be a need to “deprive” the world of rare species images and information due to threats of poaching etc. But of course not posting on the Internet is anyone’s right and choice, for any reason.

3 Likes

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.