On the importance of raising awareness of properly tagging observations of cultivated plants in Europe

iNaturalist is undoubtedly a site with immense potential.
Unfortunately, many users do not take care to read and put into practice the recommendations contained in the site instructions. Thus, many observations of cultivated plants are posted untagged.
Many of these users are simply new and have still not understood that iNaturalist is not a gardening website, others are simply superficial.
I believe we all could spend some time raising awareness among users who post pictures of cultivated plants without the proper tag.
I have already started doing this with the observations from Italy. Of course I cannot deal with all observations from Europe on my own.
I propose you to use a standard form to be translated in various European languages.

Publicado el octubre 15, 2018 05:03 TARDE por italopithecus italopithecus


I totally agree with your viewpoint, this is not a gardening website and this should be highlighted on the home page more clearly. I am sometimes flabbergasted by some of the observations that are posted. A plant in a pot on your balcony is not acceptable! I think the final option should be a flagged for deletion or inappropriate material option to be operated by approved verificators as these records are just cluttering up the system and demeaning the real and valuable records and the value of the website itself.
I would also like to make one comment about member profiles. John is a Naturalist does not mean anything. There are always novices, amateurs, educated, experienced and professional members so why not distinguish this and either create a verificator hierarchy or state who and what you are.

Publicado por bobwardell hace más de 5 años

thanks @bobwardell

Publicado por italopithecus hace más de 5 años

The distribution of cultivated plants is useful for science, so I do not advocate deleting these observations. However, they should be clearly annotated as cultivated, though the line is often blurred.
I suspect that some form of artificial intelligent algorithm could easily be trained to identify suspect observations.

Publicado por qgroom hace más de 5 años

I agree totally. I'll keep a closer eye on what is posted through Romania for cultivated/potted plants. I think the first, and already implemented step, is to tag the organism as "not wild" if it is obviously in a garden/park/pot.

Publicado por conradaltmann hace más de 5 años

I agree that it should be easy to tag observations as "not wild". The amount of observed cultivated roses, for example, takes quite some time: Check image and locality, scroll down, tag.

Publicado por torsten hace más de 5 años

I do my best marking cultivated plants.

Publicado por annemirdl hace más de 5 años

There is one method observations are autotagged as cultivated by iNat.:

"The system will vote that the observation is not wild/naturalized if there are at least 10 other observations of a genus or lower in the smallest county-, state-, or country-equivalent place that contains this observation and 80% or more of those observations have been marked as not wild/naturalized."

Publicado por bouteloua hace más de 5 años

I will try and make more effort not to simply ignore these observations! If it's in a pot, it's usually obvious, but the line is awfully blurry beyond that point...

Publicado por jeremybarker hace más de 5 años

A global issue, not just European. In SE Asia, where most of my attention is directed, observations posted from Singapore, Malaysia and the Philippines are particularly rife with unmarked observations of cultivated plants. Much of this, at least in SE Asia, seems to be due to a lack of clear communication of the intended function of iNat from the outset to the end user. A dialogue from earlier today to illustrate the lack of awareness: https://www.inaturalist.org/observations/10200929
Along the lines of @bobwardell 's suggestion above, perhaps some sort of an iNat 'mission statement' can clarify and communicate this to all new account creators in a way that is sensitive to the diversity of interests that various users might have, and also in a way that still encourages use of the platform.

Publicado por shawnodonnell hace más de 5 años

I would consider the deletion only as an "extrema ratio" in the case of observations that are obviously unacceptable.
Despite being clearly cultivated, plants grown in green spaces could be acceptable only if tagged since they could represent a potential source of escapees. On the other hand, observations of plants in pots, in my opinion, should be maintained only if (of course) tagged and well representative of the typical morphology of a species. As regards there are many observations showing just a flower of a cultivated plant. In these cases I wonder which is the usefulness of these latter observations.
I think that we all could hope to have to do with mature users with whom it is possible to discuss freely and to try to make them understand that every community has its rules and everyone should behave they way it is recommended.

I wonder if users (or most of them) really read and understand the site instructions that are sent when someone register to the site. Maybe a translation in various languages could be useful to make users understand the site policy? In the case, I would be available to translate them in Italian.

Regarding marking observations as "not wild", I have faced some cases of distinctly cultivated plants where some users had (preventively?) marked the "wild" option. In these case only at least two votes for "not wild" can guarantee that the observation would be placed among the casual ones.

In the end, I thank you all for the comments and especially those who have made the effort of tagging this kind of observations so far. Maybe we could check if a more or less vast area of Europe is still unguarded and produces many untagged observations of cultivated plants.

Publicado por italopithecus hace más de 5 años

AI tagging would certainly be a great idea. This will probably generate a few mistake, but eventually will leave substancially less work to correct that non tagging would do to ID properly.

Problem is mainly due to, imho, the mobile app - very simplified from the web version, thus far more accessible. Very usefull, with far more people - I use this with my own pupils - but on side aspect is that it drains in the community people not concerned by the project. I don't blame them, iNat is a very usefull tool, and this should go on. But this generate problems like this that we'll have to deal with.
Many people - if not most - using the mobile app (I mean, ONLY the mobile app) simply don't understand where the ID is coming from, and what iNaturalist is made for. They simply want a quick ID of what they watched.

Though generating problems to be solved, this type of users is however good to the community, I think. They generate data, some getting this way more involved into nature understanding and conservation, and can eventually become regular users, some simply watching very interesting obs. without even knowing about that (I must check, but I think I remember the first obs. of Spilostethus from US was made on the mobile app., if I don't mistake, for example...).
In many, if not most cases, warning could be useless, because some type of users simply don't take it in consideration, simply ponctually looking for a quick ID. Warning should be kept on, though, as sometimes usefull, but AI automatised tagging would be a good idea, as a beginning. What ever would be the ID algorithm, it could be improved later if causing trouble, but the principle remains a very good idea, I think.

Publicado por fabienpiednoir hace más de 5 años

@fabienpiednoir thanks for your contribution. I do not use iNat as a mobile app and I did not know these aspects you cited.
I agree with you that "casual" users may represent most of the sources of this problem.
I am also in agreement with you that in many cases warning and recommendations are of no use since these casual users just upload some photos on the app and then disappear. The problem is that they are many and they create a huge workload if someone would like to tag all the observations that are untagged.
I would be ok for the use of the algorithm for the automatic tagging even though I am a little bit concerned for the fact that I could tag as cultivated the observations of ornamental plants that, instead, are sometimes found escaped from cultivation.

Publicado por italopithecus hace más de 5 años

The algorithm suggested by @bouteloua would be just a start; it could be improved with time, e. g. by checking a local list of known escaped / naturalised species.
Plus, those species would still be to be IDed : identifiers could further recognise them as "wild", and manually remove the automatical "non-wild" tag. Of course, one of the main condition of the bot would be not to overpass a human user, i. e. not to remove a manually "wild" tagged obs, only to tag some of the untagged with a "non-wild" tag...

Think you should download and try the app - at least to see how it works, this is instructive, to understand where many of the obs. you check do come from; even for experimented users, it can be of some use, as, though very simplified, it automaticallly add GPS coordinate and date : for example, I use it for species mapping : just looking to specimens, not to care about date or place, and photograph them. The app will then use the automatically generated time/place data of the photo to create the obs., just have to add the name (if known...)

Another great thing for identifiers would be the mobile app-made obs. to be notified more clearly (it is currently, but not in a very visible way, I think...)

Publicado por fabienpiednoir hace más de 5 años

I often post this message in those observations made by users who have posted many (more than 10) observations of cultivated plants that are still untagged:

"Observations of cultivated plants and animals in captivity are welcome here, but take into account that this is not a gardening website and in these cases it is requested to correctly tag these observations checking the specific box. This is the only way users that can filter observations in order to see only those of wild organisms.
Maybe you could take a look at the website instructions before posting further observations:
You are also kindly invited to correctly tag the observations of cultivated plants you have posted so far.

You could evaluate to use the same message in your language.
Some users have demonstrated to have understood thet they were behaving unproperly but widespread efficacy cannot be expected.

Publicado por italopithecus hace más de 5 años

Totally agree! Though sometimes it is hard to tell - for instance in grassfield, gardens, etc. When I am travelling I take also pictures of grasses and weeds for instance.

Publicado por fero hace más de 5 años

Today was a very bad day for plants in pots, cut flowers in vases, plants on balconies, street trees, Yuccas in every garden in northern Portugal and even pomegranates!

Publicado por bobwardell hace más de 5 años
Publicado por italopithecus hace más de 5 años
Publicado por italopithecus hace más de 5 años

There has to be a delete option for this type of ………….!

Publicado por bobwardell hace más de 5 años

The mobile app... I recently found some of mines with picture of a mug of coffee... Was doing a demo to a colleague, didn't realize it took the photo... If I can make mistakes of my own, I imagine newbies... hopefully, @bouteloua is so benevolent and patient with beginners...

Publicado por fabienpiednoir hace más de 5 años

Could AI used in Inat automatically refuse such stupidities as toilet paper or human faces, or other stuff unrelated to Inat?

Publicado por fero hace más de 5 años

Considering the mistakes done by the AI before, I don't think this would be a great ID to automatically refuse; but, I imagine it could be improve by considering critters such as using the mobile app, AI recognition of human faces or other identified unproper subject, etc., to send a warning to the user and require an confirmation - this would be enough to stop, if not all, most of the test / mistakes.

Maybe we can imagine in such cases the AI could send a warning to a moderator too...

Publicado por fabienpiednoir hace más de 5 años

On the subject of human faces - I was wondering what the justification of having Homo sapiens in the database anyway was? and obviously the next time I photograph a Homo erectus or a Homo neanderthalensis walking down the street I will surely add my record to iNat! LOL!

Publicado por bobwardell hace más de 5 años

One more comment on the subject of human faces. A colleague mentioned the problem when students doing projects involving iNat observations just uploaded images found on the web in order to finish their assignments. In that discussion it was (jokingly?) suggested that if the student was required to be in the picture together with the subject, faking an observation would be more difficult.

Since iNat is more and more commonly used as an education tool, and teachers can easily come up with things like that, we might be getting lots of faces or other assignment identifiers in the observations. Irrelevant to all but the teachers.

Publicado por torsten hace más de 5 años

Well, though spotting H. sapiens is clearly not the first aim on iNat, as the database is automatically incremented by external names providers, then it is necessary in the database; as Cassi noticed before, most of human faces spotting are the result of trials, mistakes, etc. rather than true malevolence;
therefore, simply properly identifying them as H. sapiens makes unnecessary tagging as spam.
The algorithm integrates H. sapiens ID in the research level score : being tagged as H. sapiens makes it exactly the same as non wild, no direct evidence, etc.

So, most of time, this is not really a problem, except when a picture of another person is posted (and often badly used...) against his/her will...

I saw once various obs. of what appeared to be students' obs. , figuring who was manifestlyanother student, or even the teacher himself, and tagging him badly, as, e.g., Toad, Badger, etc.
I suppose teacher using the app with their student then watch over their work... I, personnaly, do. This personnaly never happened to me, but I suppose my own student are aware, or, at least, fearfull, of what could happen to them if I ever discovered a tagging of another student, or myself, as an insulting animal... Teachers leading students and guiding them on iNat should be able to manage them, I think. And I can be pretty imaginative in case of such dysfunctions... :D

Publicado por fabienpiednoir hace más de 5 años

Still - extinct species from the Pleistocene +/- 10,000 years are not observations of living things. - I just checked - Woolly Mammoth is in the database with Mr. Neanderthal!

Publicado por bobwardell hace más de 5 años

After all, I think that both teachers and parents should be responsible for what their students/children do on iNaturalist.
It is good that on the web there is something like iNat that is useful to spread the knowledge on natural sciences, but this should not allow kids to do everything on iNat as well as adults to let kids do everything on iNat.

Publicado por italopithecus hace más de 5 años

a local endemics... neither so young...

Publicado por italopithecus hace más de 5 años

How to create a project without informing how to behave on iNat and its consequences:

more than 50000 observations, mostly of cultivated plants and captive animals, practically none has reached research grade. What is the meaning of these initiatives?

Publicado por italopithecus hace más de 5 años

Hi blue, you're right.
What can I do besides this done by myself?
Good evening.

Publicado por naturalista1989 hace alrededor de 4 años

Agregar un comentario

Acceder o Crear una cuenta para agregar comentarios.