Auto-suggest: A love/hate relationship that causes me to ramble

Before I left for the Del Rio BioBlitz I decided to use my camera less and get back to using my cell phone and the iNat app for more of the observations that I'm not trying to turn into a photography show piece. (Sarcasm there, friends.) That was sort of a trial-run for the City Nature Challenge, where I needed to still get a lot of observations in, but without sitting at the laptop editing like mad. It has worked wonderfully! I no longer sit for hours reviewing, cropping, and being overly particular about every image. As soon as I leave a location I hit UPLOAD, and then I only have a few distance shots of birds and such to add at home. And even better, I already have IDs and comments to review when I finally DO sit down at the laptop.

I've come to realize a couple of things though, none of which are terribly surprising. 1) I love having ID suggestions so easily at my fingertips. 2) I still suck at picking the right ID.

Auto-suggest is great for very common species in very common iNat areas with clear pictures. How it can tell the difference between some of the Odonates is just mind blowing to me. But even in DFW, I frequently pick the wrong thing from Auto-suggest if I have no idea. It doesn't help that the top suggestions are occasionally species that don't occur in DFW at all.

I've debated about using a more conservative selection, but I wonder what new users are experiencing, so I continue to explore it. (I'm one of those in the "best guess" camp that tries to put my best guess on the observation and revise or withdraw as appropriate. Yes, there are many people that despise this approach, and I'm ok with that because they have their own camp. It takes all kinds...)

I've also noticed some glitchy user errors on the website auto-suggest when I'm uploading new observations. For example, I might know the tree is an oak, but if I'm not sure WHICH oak, I scroll through using the images for comparison. Often, I'll go back and select the genus or family thinking I'll dig into it more later, only to find that somehow I unintentionally selected the one oak species NOT in DFW. ?!?! I've tagged a couple of recent ones that I recall.

One thing I've come to appreciate more now that I'm using the auto-suggest is the quality and type of photos on the taxon page, that is displayed on the auto-suggest. If I know I have a Packera species (and that in itself is progress!) but not which one, I'll scroll through the suggestions to compare leaf images. Except there aren't any. There are eleventy-six pictures of the flowers (which all look the same) but not a single leaf picture. When I come across this situation (at the laptop) I will go through the images and swap out the first few to have significant identifying features pop up first. A couple of flower pics, but also leaves, whole plant, seeds, fruit, bark, etc.

I hope, despite all the incorrect IDs I've put out there --with or without the auto-suggest--, that I'm learning every day and somehow still helping to improve iNat in whatever little way I can.

由使用者 kimberlietx kimberlietx2018年05月11日 22:58 所貼文






4月 29, 2018 11:54 CDT



瘦姬蜂屬 ( Ophion)




4月 28, 2018 21:58 CDT





5月 9, 2018 11:26 CDT


I am loving the auto suggest feature, too, and am happiest that it gives me a place to start when trying to figure out an ID. I had never used the phone app until I had to learn it to teach it in prep for this year's Nature Challenge. Wow! So easy, and having the IDs pop up instantly opened up a whole new use for iNat...field guide for everything (plants, bugs, trees) easy to carry while hiking. Birds and far-off mammals still require the camera and ID at home, but I'm happy to have the ability to use the app in the field, too.

發佈由 naturemom 約 6 年 前

THANK YOU for helping swap out identifiable leaf/etc pics on taxon pages. It helps so much!

發佈由 bouteloua 約 6 年 前

Like you, I also have a love/hate relationship with the auto-suggestions. I’m amazed when it gets the species ID right, but I’m also pretty frustrated when it gives a suggestion of something only found in New Zealand. I’ve found that it’s popping up lots of mis-ID’s on new users’ observations, so I’m having to become a bit more diligent on ‘double checking’ the observations made in our specific area.

It’s also only as good as what’s been documented where you are. In DFW, it’s pretty freaking awesome, but it totally was lacking in Del Rio, and on the plants and bugs when I went to the Lower Rio Grande Valley... It’s ok — those areas just demand a few more trips or growing iNat communities that ‘double check’ the ID’s on stuff too! :)

Something that I keep reminding myself is that it’s not perfect yet — and that with each and every observation and ID that we do, it gets better — just like iNat in general. :)

發佈由 sambiology 約 6 年 前

Thanks for writing this - so true! I go back and forth from using my phone and using my camera. We usually settle for my husband using his phone while I use the camera. So then we can try to identify stuff in the field, plus (maybe) have better pictures to look at when we get home. Of course, when we get home he's made 40 observations and is happily checking to see his confirmations, while I have 400 pictures to go through and several hours of work before I can post anything! Incidentally, I JUST learned how to use the geotagging app a couple of weeks ago, which almost made me cry with joy! Before, I was spending many minutes manually putting in coordinates for each observation! =^_^= Oh, well, at least now now it seems practically fast!

發佈由 lisa281 大約 6 年 前


登入註冊 添加評論