Asteroid Zoo Talk

Discard sets once nth agreement achieved?

  • Almohan by Almohan

    I often find that on any given set, whether I found an asteroid or the set proves defective, I often am not the only one to mention it in the observation field, so can't these sets be put aside once a certain number of comments have agreed on them? If not, it may prove a bit tedious to be the nth person making the same comment after other people. Or maybe we could access the comments (if any) before getting to examine the set? That would mean changing the interface and I have no idea how difficult that'd be, but I guess it would suppress superfluous viewings . . .

    Posted

  • CTidwell3 by CTidwell3

    They definitely can't modify it to let you see comments before you mark things, because that would completely change the 'unbiased' nature of the results. For example, a set with a hard to notice asteroid (1 out of 10 users seeing it) would suddenly jump to 8-10 seeing it when they jump to the comments, see that there is an asteroid, then all have an idea where to look to mark it. The goal with these crowd-sourced projects is to have the marking be blind and once submitted, uneditable, since the lack of marking something can provide just as much information as marking something at times.

    They have mentioned that we are still in the phase of refining just how many people need to see an image before the set is considered finished and that they initially have this number of views higher so that they can create a better statistical data set to determine the optimal number of people that need to view any set.

    http://talk.asteroidzoo.org/#/boards/BAZ0000002/discussions/DAZ00000kq

    That link is similar to the one I am remembering, with the answer by Dr. Science (#3) at the bottom of page 1 being what I am talking about. If you need more than 2 people two people to identify an asteroid, then how many views of that set need submitting to make sure that number of identifications is seen. Maybe its 6 submission. Maybe its 10. Maybe its 12. Intiailly it will probably be highter than needed, but the results will let them fine-tune how many times you see a set.

    I console myself with when I see comments on images sets I have just marked that I am getting that set one more submission closer to being removed from rotation.

    Posted

  • Almohan by Almohan

    Thanks for caring to answer, you are making quite a point there (and a few more at that).
    Now where's that rollingupsleeves smiley again?

    Posted

  • Almohan by Almohan

    Maybe one more suggestion (though I'd think someone else already pointed that out somewhere): when I want to mention that a set is defective, I have to go "nothing" first, which is not a logical way do it. There should definitely be an option allowing you to quit tagging the set and just label it as useless, don't you think?

    Posted

  • a349 by a349

    I thought that the aim is to find as many moving objects (Asteroids, NEOs, TNOs etc ..) and nothing else.
    If you want to study our efficiency, they should inject artificial moving objects in various magnitudes and high quality
    frames. Then they can calculate our efficiency in detecting them.

    Posted

  • Almohan by Almohan

    Well I was just thinking ergonomy, I don't pretend to have anything to teach the people who run this thing, I was just talking as a mere occasional user -- one who's incredibly thrilled when he spots an object and then finds out that no-one seems to have before him since the relevent Talk section is empty, though I don't always get the "Guess what! You're the first to see the set!" message in green.
    I was just thinking that some improvements could be made, but of course I have no idea what pain it would take the people in charge.

    Posted