That seems like a good UI choice if this visual representation of CV confidence makes it to the desktop (web) version of iNaturalist. But “hover” is a state that only exists when there’s some sort of pointer/cursor, not with a standard touch screen as on phones. Am I missing something about your suggestion?
I’m using Next for the first time today, and I agree I think that the magnifying glass button is too out of the way. On the suggestions page I would prefer it large and easier to access. Since I observe a lot of bugs and tricky plants, the CV suggestions often aren’t very useful. It’s a long way to move my finger all the way to the top right of the screen!
I’d also like to have it on the main observation prep page, probably to the left of the big green “Add an ID” button. There are many cases where I know the CV won’t be reliable, so that would save an extra click. (Edit: this proposal would make this unnecesary)
I like that there are now easy options for “Plant”, “Insect”, “Arachnid”, etc. However I rarely actually identify to those levels; going down a couple steps to Dicots, Pterygota, Spiders, or a specific family saves broad identifiers some work and makes it a bit more likely that my observation will fall into the ID filter of identifiers who are more likely to know the species.
3 Likes
Ooh I would love to be able to see a tree and choose the common ancestor of two or three of the suggestions, I basically do that manually regularly. That seems too complex to be implemented for general users though.
4 Likes
Two thirds of GSB obs were plants. That is beyond not helpful. The ‘trivial’ third can be usefully divided to reach the right identifier.
Monophyletic iconic taxa are more difficult for plants, apart from a few obvious ones (such as mosses, ferns).
For one thing, people not very familiar with plant taxonomy won’t know what a “gymnosperm” or an “angiosperm” is, nor will they know “monocots” and “dicots”. People learn to differentiate plants differently (“tree”, “shrub”, “herbaceous plant”, “flower”). So there will probably be a lot more misidentifications and observations left as “unknown” if people don’t have the option to select “plant”. And if an identifier disagrees with them, the observation will be kicked back to Kingdom: Plantae anyway.
For another thing, plants, unlike animals, do not have a determined morphology (where a new leaf or branch will grow is largely determined by phytohormone responses to environmental conditions and not cell lineage)
And different orders and families within the same class look significantly different. (Will non-botany folks recognise a non-flowering cactus as “Magnoliopsida”?) Iconic animal taxa have at least some easy-to-see features conserved in almost all species. (number of appendages, hair, feathers, scales, exoskeleton, etc.)
Most people think of all that green stuff as just “plants”. If we want this problem to be fixed we need to start outside iNat: educate more people on botany to fix their plant-blindness, teach them basic plant taxonomy, etc.
I don’t think, making usability worse for other users because it may be more fair/equal is a particularly good way to fixing this.
5 Likes
is there a reason a community science site has to be strictly monophylletic, even for things like iconic taxa that are meant for the newest amateurs? It seems like imperfect iconic taxa like ‘conifers’ do more good than harm. Otherwise the site just ends up coming off like it emphasizes plants less than animals. As we learn more about evolution and how much things intergrade the concept of strict monophylly seems questionably useful anyway.
I say just go with it.
2 Likes
I‘d be against it as it would require non-monophyletic taxa in the iNat taxonomy. IMO as a citizen-science site it should „follow science“ and the best way to classify organisms currently is cladistics/phylogeny that works with species and monophyla of different sizes and ages. (I currently have a draft for a topic about roughly this that I started writing yesterday)
There could be a solution with better (and searchable) paraphyletic placeholders/tags, but that would require iNat to change how that whole system works which I think is unlikely.
Why do you think that strict monophyly is questionably useful?
Just a quick reminder that the topics in this section of the forum are specifically focused on iNat Next and in this case, implementation of the CV in the new app. Please keep the conversation focused on topics that are linked to the new app to make it easier for outcomes for the app to be surfaced and used.
If conversation about taxonomy in general is desired, we can reopen one of the older threads about this, like
https://forum.inaturalist.org/t/should-new-disruptive-technologies-be-used-for-classification-in-ancient-linnaean-rankings/50183
and move posts there.
1 Like
“follow science” isn’t meaningful in this context though, it’s not inherently unscientific to group organisms based on commonalities that are not genetic lineage.
because i don’t think it’s attainable in a meaningful way. I think things mix and hybridize, different types of genetic material (mitochondrian vs not, etc) act differently, and evolutionary convergence makes things that are not very related act similar. Plus changes happen in spurts and drawing it out like something where everyone evolved the same amount makes no sense either. A true cladistics would have aspects of a braided river rather than a truly dendric one, which means trying to be strictly monophylletic won’t be meaningful. I think one of the reasons the current attempt to classify taxa to great detail with genetics is spouting out unusable gibberish is that it turns out life and evolution are much messier than we thought and species are just never discrete entries.
Relatedly trying to keep every thread on this forum ‘on topic’ is pretty impossible and meaningless in the same sort of way and ends up leading to what feels like arbitrary censorship. But if you want to move this to that other thread you can. Since it’s not open now i can’t do it yet.
1 Like
Just made an observation where Next was “not confident” but literally all of the suggestions were beetles, most of them in the same family. The app should definitely be offering family or order here, instead of all these species/genus options
9 Likes
iNaturalist has an inordinate fondness for beetles.
3 Likes
I wanted to reply with just “Ha!” but Discourse wanted more characters so I had to write this meta-comment.
1 Like
Agreed. This is still concerning to me, and even if it somehow doesn’t cause more miss identifications. I want the ability for my self to be able to pick higher taxa quickly when uploading. I have no idea what family or subfamily the top suggestions are.
This also extends to other instances where the CV will suggest mosquito and chironomid species, but it will not show “we’re pretty sure its in x” if the taxa level is above family. At least It seems.
I’ve seen many instances where the CV could’ve suggested Nematocera, or Chironomoidea.
4 Likes
all platforms and all confidence ratings should have this feature
There’s an old feature request for this here that’s been under review for at least a couple years now, I’m very curious what’s impeding it.
6 Likes
This topic was automatically closed after 60 days. New replies are no longer allowed.