There are a few projects named “Unknown / a taxonomic group”, and observations are added to such projects. Strangely, the observation has no identification which would fit to that group at all, only far coarser ids.
E.g. observation
https://www.inaturalist.org/observations/119407058
has the identifications of “Plantae” and “Magnoliopsida”, but it is in the project “Unknown / Amygdaloideae”.
That is, there are identifiers browsing observations at current coarse taxonomic rank, actually do an identification at an intermediate rank, and add the observation to such a project - without providing an identification in the field for the identification.
And thus, the observation stays at the coarse rank. And specialists for that group likely won’t see it.
Why? Just why?
I’ve seen it too and wondered the same thing. I just go ahead and look at the actual IDs, but it would be interesting to understand the rationale.
I can’t give you all the background, but @jeanphilippeb is responsible for creating the “Unknown” projects. I believe they are sorted into those by the CV. However, they are not given IDs, they are still Unknowns, so identifiers will still see them. Real humans (like me) can then go through Unknown projects and give the observations correct IDs.
So it’s all done automatically by the CV?
Hello,
Some background information on “Unknown / *” projects is available here : https://www.inaturalist.org/journal/jeanphilippeb/73398-draft-for-creating-projects-for-unknown-observations
Hopefully either Jeanphiippeb or @dianastuder will explain. (I think it was Diana who suggested it originally?) It is all about helping the huge pile of “Unknowns” get identified.
Thanks! I hadn’t read that before.
Thanks for the hints. So the process is automated. That’s why no identification is provided.
No the Pre-Mavericks were my idea. (I am clearing the last / latest 350 African Pre-Mavericks)
These yellow label projects are JP helping to move Unknowns along.
Use his suggestions as a starting point, which may agree or disagree with your own ID.
I think it is … and yellow label suggests a broader taxon which covers mine - tick.
No idea, but yellow label suggests - I can consider those.
Or I can @mention a trusted taxon identifier - if it looks reasonable, but I have no idea from cheese.
It is another tool, to use or ignore, as you see fit. (But frankly a more useful option than just dump it in dicots - about 500 notifications yesterday!) Dicots to annotate as fruit, or add to the tree project - that moves the ID forward.
I will not deliberately use a wrong ID, but if I try it and it turns out to be wrong (it WAS seen by a taxon specialist) - I will withdraw with my notification.
@bernhard_hiller there is no ID added, as the suggestions are automated. They may also be wrong. Use your own discretion.
There is a link to JP’s journal post in an earlier comment.
If you are a taxon specialist - you can start your IDs from his yellow label project for your taxon - easier for you to move them to a finer ID, or out of that wrong suggestion.
Increases in specificity of taxon moves IDs along.
I just tested two of those “Unknown/*” projects in connection with the “Mission Impossible” task. Getting from the project to the ID module takes some effort. Then too many cultivated plants showed up, and also multi-species observations which were already marked as “id is as good as possible”; but there were some observations marked cultivated which to me seemed wild.
After adjusting for that, I could confirm many observations at the familiy level, a few belonged to other families, and in several cases I could not decide. But there were not many observations in Unknown/Asteraceae - because it does not include the obsverations which could be guessed to subfamily level. So open the next project, adjust the URL as before, and iterate over them.
After all, it seems to be a useful tool. Since the projects were generated automatically, it could be possible to adjust them automatically for future versions of the “Mission Impossible” task.
You can also start from your preferred ID link. Then use the project filter to add the relevant project name.
No, that would not work, as there I take a look at “everything”, including captive, missing photo/sound, already RG, already reviewd, etc. So I can help beginners when they made a little mistake, find wrong RGs, and also learn from others. And I see my own observations, too, and can detect issues with them (sort order of photos, wrong combinations, …).
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.