Snapshot Serengeti Talk

Does this forum have a search feature?

  • Sylverone by Sylverone

    Somehow I've been unable to find one. It seems strange for it to be missing, since I think all of the other Zooniverse forums or talk areas have a search function.

    Posted

  • dms246 by dms246 moderator

    You're not mistaken, @Sylverone - we don't yet have a search feature in the forums here. We've been told the Zooniverse guys are working on it, though. 😃

    Posted

  • Sylverone by Sylverone

    Good to know. I wonder if there will still be any data left to classify by then? 😃 Is this simply a very popular project, or is the dataset also smaller than some other Zooniverse projects?

    Posted

  • kosmala by kosmala scientist in response to Sylverone's comment.

    I think part of the reason these data sets go so fast here is that we don't require too many people to look at each image. I've been told that Galaxy Zoo, for example, asks for 60 people to look at each galaxy. We don't feel we need quite that level of scrutiny because, well, once 5 people have said there's 1 giraffe in an image, we don't really need another 55 people telling us there is 1 giraffe in that image. I believe we can do a good job with identification with 5-25 people looking at each image, depending on how much agreement there is. So that, and also, yes, I think it's rather popular. (And Arfon said that the Z-team is working on hashtags -- and therefore search -- this week.)

    Posted

  • dms246 by dms246 moderator

    I suspect this is also a very popular project, because who doesn't enjoy looking at photos of beautiful wild animals in their natural habitat? And there's the "immediate gratification" factor when we find an image of a cheetah, or a cute baby zebra. 😃 You don't get that in quite the same way in many of the other projects. It's more concrete than many of the other Zooniverse projects, too, and we all start off with much of the knowledge needed - we all know an elephant when we see one, and can tell the difference between a giraffe and a lion. So the learning curve is gentler compared to other projects. That also means fewer classifications per image are needed, and it's only the hard to make out images that require more classifications to produce reliable data.

    Posted

  • Sylverone by Sylverone

    That makes sense. Thanks for the replies. Upon finding the project (after all the initial data had been gone through) I could immediately see why it would be popular, but it wasn't quite clear why the images were burned through so quickly. Fewer required identifications explains it, especially when considering how many of the images are just waving grass. I'm sure classifiers go through those like nobody's business. 😃

    Posted