Plankton Portal Talk

Worried

  • mlmuniz by mlmuniz

    I am learning a great deal, and enjoying this very much, but I am sometimes concerned that I might mislabel something. For example, I read a post that helped me see that I was labeling young rockets as thalassos, since the tails are not exactly prominent. How concerned should I be that I might have labeled these or something else incorrectly?

    Posted

  • jessicaluo by jessicaluo scientist, admin

    Hi @mlmuniz, Don't worry about misclassifying things! There are going to be multiple volunteers looking at each image so we can get a consensus from people on what animal is in the image.

    Also, we know that almost everyone misclassifies things as thalasso, just because they appear to be round and generic looking in the classification scheme, so protists or other small things often get labelled as thalasso. In our data cleanup, we will probably end up flagging all thalasso classifications that are below a certain size threshold because we know that thalasso are not the size of protists. We also know that people generally confuse doliolids for thalasso, or as you said, young rockets for thalasso. With this information in hand, we can readily correct the classifications.

    This also applies to all our organisms. We have a general idea of how common or rare they are (as well as what depths we generally expect them to occur), and what the common mistakes people will make with classifications, and thus can correct for it. So don't worry about it -- any information you can give us is more than what we already have, so we're grateful for it!

    Hopefully that assuages your fears! We will also make a blog post about these issues. Many people have brought it up.

    Thanks for all your help with this project!

    Jessica

    Posted

  • mlmuniz by mlmuniz in response to jessicaluo's comment.

    Jessica,

    Thank you. I do feel better now.

    I love all the Zooniverse projects, but it seems I keep coming back to this one. It's really quite interesting. I'm a fan of all things ocean, and am surprised at how much these little critters have intrigued me. My expectation has always been that I would be more taken with fish or large crustaceans. With the exception of a few larger ocean animals, this area seems to have become one of my new favorites.

    I'm a secondary public education teacher whose discipline is English. Oddly, I have a nagging love for the sciences and have already shared SETI and Zooniverse with my science coworkers. I believe this is a great opportunity for students to expand their knowledge of all things science as well as explore possible career opportunities.

    Personally, however, I simply appreciate being allowed to participate.

    Thanks again.

    mlmuniz

    Posted

  • Forestchick by Forestchick

    The user interface is very confusing! I thought I was doing the tutorial and was just messing around with it! But it turned out not to be the tutorial anymore. So now there is a bunch of data in my profile that is inaccurate. Is there a way to delete those false entry's? Also you need to tell people that this does not work well on iPads. Also there should be an undo after you have missmarked something., but that may just be an iPad problem?

    Posted

  • astroboyOW by astroboyOW

    I've just noticed that, when I finish with an image and it slides up and out of view, a portion of the image has been hidden behind Planton Portal's data and buttons at the bottom. I can't find a way in my browser to fix this. Is there a way around it?

    Thanks,
    Jeff

    Posted

  • yshish by yshish moderator, translator in response to astroboyOW's comment.

    hm, weird, sounds as a bug to me. could you tell us what OS and browser you use?
    i'll tell @jessicaluo about your problem. wait for the answer.
    thanks for letting us know.

    Posted

  • jessicaluo by jessicaluo scientist, admin

    Thanks for letting us know! We will check out this bug.

    Posted

  • astroboyOW by astroboyOW

    Win7 and Firefox. Actually, there are a couple of other bugs. I have scroll up and down slightly, even to see the apparently truncated image that I am getting, and drawing arrows gets really funky near the top and bottom edges. The PP menu bar intrudes if I try to draw close to the top, and I can extend lines down into the data/button area at the bottom. The line seems to go "behind" the bar in the latter case, but I can continue to draw and finish the line.

    Posted

  • yshish by yshish moderator, translator in response to astroboyOW's comment.

    and have you tried to zoom out in your browser? it could help to solve the problem with tagging plankton close to the top edge, because the top menu won't be popping up to the image.

    Posted

  • bumishness by bumishness admin

    Sorry to hear about your troubles astroboyOW. I'll be happy to help try and get to the bottom of this.

    Could you take a screenshot of the behavior you are seeing? I think I understand what the problem is, but I'd like to confirm before I go troubleshooting.

    Posted

  • astroboyOW by astroboyOW

    yshish, I did try zooming out. All it did was make the text and icons at the bottom smaller, without reducing the height of the bar.

    bumishness, I couldn't get a screenshot of the behavior at the top of the screen, but here is one of the bottom:

    http://s747.photobucket.com/user/astroboypb/media/PPscreenshot-bottom.jpg.html?filters[user]=105274013&filters[recent]=1&sort=1&o=0

    The cursor doesn't show up, but on my screen it's there, extended into the bar where the line would go. I can move it around, go above the bar and back below it, with the line following it, just as though the bar isn't there.

    Posted

  • astroboyOW by astroboyOW

    Since I'm evidently missing a significant portion of each image, should I not do any more classifying until this issue is resolved?

    Posted

  • Bram_Groeneveld by Bram_Groeneveld

    Hi, I really don't understand the tutorial, but would really like to contribute to this project. Could someone help me out please? What's with the axes and lines?

    Posted

  • yshish by yshish moderator, translator

    hi @Bram

    i'm here to help you, just need to know what exactly is a problem for you. in the tutorial, you get an image full of creatures. we want you to measure how the creature is long and wide at the maximum. the 1st axis is always a length and the 2nd one is a width.

    when you put these axes wrong, the tutorial makes a guide lines for you and ask you to correct them.

    hope, this is helpful. if not, let me know:)

    Zuzana

    Posted

  • SPToad by SPToad in response to jessicaluo's comment.

    So, because you plan to tag the smaller thalassos, would it be helpful if we lean toward labeling uncertain organisms as thalassos? So that you would be more likely to check it and correct it if we misidentified it? I'm having the same ID problems as mlmuniz described above. I keep wanting a tag that says, "Unknown round thing - please check." Sometimes it's just blurry or a weird angle, and I don't want to skip it, because it's definitely there.

    Posted

  • DZM by DZM admin

    Hi @SPToad, there's actually a really good reason that Zooniverse projects never have an "unknown" button of any sort. Remember that many people see each image, so at the end of the day, each image will end up with many classifications. We've found repeatedly that the wisdom of crowds here applies... the best guesses of 25 people, even on an unclear object, will more often than not produce the same answer that a professional scientist would give! So we want to encourage people to guess, no matter how wildly they have to.

    Additionally, if those 25 people end up classifying an image in 15 different ways, and there's no consensus at all, that actually serves the exact same purpose of "unknown, please check." An object that ends up with a lot of wildly disparate guesses will automatically be flagged for review after the data reduction.

    In other words: a wild guess is far more helpful to the scientists than marking nothing or marking "something here but I don't know what." 😃

    Hope this helps!!

    Posted

  • yshish by yshish moderator, translator in response to SPToad's comment.

    Hi SPToad,

    I'm not a scientist but I'd like to reply as well.

    However I don't think that choosing a thalasso for every round unclear creature is the best choice since you know that they're rare and appear bigger, I agree that it's always better to guess something (even a thalasso) than mark nothing at all.

    If you were interested in Thalasso's appearing here are a few images of them with the typical and usually visible drawing on the body:

    [1st], [2nd], [3rd], [4th]

    ..or you can go to the Search and look for a #thalasso by yourself.

    Next time when you get such round unclear creature click on the 'discuss' button after the classification and leave a comment that you have problems with the identification. I'll do my best to help you with that so you'll be able to recognize that in future.

    Thank you for the question and keep asking when feeling lost:)

    Regards,

    Zuzi

    Posted

  • SPToad by SPToad

    Okay. I've been trying not to skip any, but it's hard to get myself to make really wild guesses, so I might have skipped a few in utter frustration. They tended to be the very, very blurry ones. Recently, though, I've found that going to do something else for a while and then coming back to it later helps.

    Posted

  • yshish by yshish moderator, translator in response to SPToad's comment.

    Thanks for trying! Just remember that you don't have to classify everything (f. e.: too blurry animals which can't be identified even by the scientists, animals which don't belong to the categories we classify like some protists, larval fishes, larval barnacles etc.) If you see something very different leave a comment there. When it is too blurry that you really can't tell, than it is better not to guess and ignore it.

    Posted

  • DZM by DZM admin

    @yshish... really? "When it is too blurry that you really can't tell, than it is better not to guess and ignore it."

    That would be different than on other projects. On Snapshot Serengeti and Chicago Wildlife Watch, for instance, users are encouraged to guess even if the animal is so far off, or so close in, or so blurry, that it's impossible to really tell. To quite a scientist over there: "if you see something, you can't mark nothing."

    Do the scientists here want something different? If so, okay! But for sure, on several other projects, wild guesses are preferred to leaving something unmarked. It produces better results in the data reduction.

    Thanks for clarifying! 😃

    Posted

  • yshish by yshish moderator, translator in response to DZM's comment.

    Hi Darren,

    Yes, I'm sure. I was told more than once not to worry about / mark the ones which are too blurry, like here: http://talk.planktonportal.org/#/subjects/APK00012u6 or here: http://talk.planktonportal.org/#/subjects/APK0006n11. I'm not able to find the thread where they told it to me but here is another one where Cedric Guigand told "Do not worry if it is too small or too blurry we just can't ID everything." and here in this thread Jessica Luo said: "Don't mark something as thalasso unless you're absolutely sure." .. Oh, and here is another thread where Cedric said: "better to leave unmarked if you are not sure".

    By the way I've just found another interesting and useful advice about marking partial organisms from Jessica Luo.

    I always try to guess when I'm not sure but sometimes you really can't tell whether it is an organism or not:] It can be just a marine snow or whatever out of the categories.

    Hopefully I'm not misleading you guys:]

    Zuzi

    Posted

  • DZM by DZM admin

    Okay, that's just a little different from how it works on some of the other projects. Now I know!

    Posted

  • Pilgrim_alas by Pilgrim_alas

    Field Guide offers more categories of plankton than are offered as possibilities for categorizing. An example is Radiolarian shown in the field guide, Round, No Tentacles group but then is not offered when you are actually classifying.

    Posted

  • yshish by yshish moderator, translator in response to Pilgrim alas's comment.

    Hi Pilgrim,

    There are actually two locations you can switch: Californian current and Mediterranean. Images of each include a bit different animals but the Field Guide includes them all! So when you classify the California current images, you won't have an option to mark Pteropods, larval FIsh or Radiolarians however you may see some occasionally.. When you classify the Mediterranean images, you won't have an option to classify Lobates, Thalasso, Medusae 4tentacles or Beroe. You should see the location where each category appears in the Field Guide as well.

    Hope it is clearer to you now 😃

    Cheers,

    Zuzi

    Posted

  • Pilgrim_alas by Pilgrim_alas

    Ok, now I understand the two image groups in the Field Guide. When I am actually classifying, there is a box at bottom that says, for example, "currently observing the California current". However, if I have an on screen image of a Radiolarian, I should ignore it ?

    Posted

  • yshish by yshish moderator, translator in response to Pilgrim alas's comment.

    Yes, you may see a pteropod/larval fish/radiolarian in California current images too, but you cannot classify them. You can leave a comment at such ones if you want to (I do it just in cases they look cool or if there is something special on them).

    Zuzi

    Posted

  • Pilgrim_alas by Pilgrim_alas in response to yshish's comment.

    Thank you

    Posted

  • jo.irisson by jo.irisson translator, scientist

    The categories we decided on are the most common/easily recognisable organisms in each ecosystem. But of course pteropods, radiolarians and others are everywhere in the oceans.

    So indeed, you should not mark them (actually you cannot...) if they are not in the list and you can tag them in Talk if you like to keep them easily accessible (but that information would not be used in quantitative studies of course).

    Posted