Weighted Identification

I am not sure if this would be helpful, I have notice from using inat that there are Experts who help iding observations. Would it be beneficial that going on these experts total correct identifications against how many they have got wrong over a certain period of time that there identification could be more valuable to help to get to a research grade id. Though I do feel it would be difficult as a lot of users do just go in and agree, maybe it could go against first iders or disagreeing iders and not just those blindly agreeing. There would have to be a way of deciding who was an expert.


That is a judgement call we make.

My ID of Dracaena was rejected … click thru to his profile, and he is studying Dracaena - then I can withdraw my own ID with confidence. Also if people are kind and take a moment to explain in comments, WHY it is or isn’t that, it helps the rest of us to learn.

This is discussed under the posts about “reputation”. It’s tricky. One approach is to create some sort of algorithm but that’s really difficult to code, intensive to run, and needs to somehow weight expertise not just volume. Otherwise inat is left somehow being the gatekeeper of who is or isn’t an expert which is iffy and problematic in a lot of ways. So it’s a hard issue.


I don’t like the idea of anyone getting more weight than any other user. Experts should be educating others on how to identify so the community can reach a consensus for an ID together, not relying on one person.


For me, the time it takes to go through the observations and add the “why” while I’m identifying has become overwhelming with the sheer number of IDs I add, so I usually just add the identification and explain if the observer asks what the difference is. I think there are some that, when asked, won’t explain, but generally, they will if asked. Ultimately, it takes a lot more time to explain, but I’m very willing to do so if it will help get the observation to the correct ID.


To add onto this, I don’t really see this being a problem currently. If experts are participating in the community, their expertise becomes clear from how they interact with the community. In my experience, they get recognized by avid users without the need for some kind of formal badge, weighting, reputation, etc.

I could see this being an issue if experts’ IDs were getting swamped out by incorrect identifications from other users, but I just don’t see that occurring (at least for the taxa I interact with regularly).


I agree with Frank. To me, Community ID works and is the best aspect of iNat. It encourages interaction between its members, and has a feeling of inclusiveness. Even an errant CV/AI identification can be rectified with a little “weight” (meaning numbers of correcting IDs), and to get that weight I have learnt to recognise who to tag for their “second opinions” based on how I see them identifying. The relatively low level of “conflict” over an ID is healthy for stimulating discussion and the passing on of expertise, even if it has to be asked for first.

The Agree buttons do change it from a single to a pair of IDs that need correcting, significantly increasing the weight required. Further, it does drop the observation from the ID pool, reducing the chance of it getting re-evaluated, but perhaps that is a “benefit” in that it encourages more experts to review the RG obs! (trying to find a silver lining here :) )


I will general explain what I was considering and looking at when I was wrong (or stuck at Genus) when asking what I missed/doing wrong to help people help me. It is easier to say an overlooked or incorrect idea than to explain why X and not Z when you have no idea what they know. Last time I recall was the color wasn’t what I was expecting, so I IDed to Genus, mentioned what I thought it was and my hesitation about it. A user swoops in, IDs to species, and comments “Adult Female” thus explaining the color difference. Asking “Why that?” would put quite a lot of explaining about patterns on the back, tail, underside of the tail, . . . and if someone tried to put that amount of work onto me, depending upon my mood, I would either ignore it, or ask why they thought the other., not that it will ever happen in the near future in regards to IDing. :)

Imagine trying to explain to someone how to successfully ID domesticated dogs in all their crazy varieties, and someone would probably be hard pressed to get genus level mismatches regularly, and higher level errors would be rare.

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