Zooniverse

 

Hi 🙂

I went to a talk on thursday by Grant Miller who works for Zooniverse, a citizen science website. He was a great speaker, and the talk was very interesting so I thought I would share some of it with you, and also address some of the criticism of this and similar projects.

If you haven’t heard of it before, the idea behind citizen science is getting the public to help sort through scientific data. This can be all kinds of things, from pictures of animals taken by motion-sensitive cameras in the Serengeti, to the different kinds of whale song. This kind of identification process can’t be done with a computer due to the imperfect nature of the images and sounds. It’s the same reason why you have to identify those squiggly letters sometimes when making a new account online—computers are just a bit thick at that kind of thing.

The reason the first Zooniverse project Galaxy Zoo was created, for example, was that the Sloan Sky Survey had recently imaged over a million galaxies, and the task of analysing them for their basic features fell to one Phd student (apparently called Kevin). He worked at this task for a whole week of 12 hour days, and managed to classify 50 000 galaxies. At this point, however, he had had more than enough, and after some words with his supervisor, probably not all savoury, and a chat with Chris Lintott, (@chrislintott) the idea for GalaxyZoo was born.

This many classifications, by the way, is called a Kevin week.

But hang on, you might be thinking. If Kevin found it that boring, why won’t we? Isn’t citizen science simply scientists making the public do the boring parts of science for them?

In a way, yes. That was the initial idea. However it quickly became much more than that. GalaxyZoo involves people answering a very simple set of questions—Is the galaxy spiral-armed or smooth? How many arms has it got? Which way is it spinning?—but soon, the people classifying them started to look beyond the central image of the galaxy. A Dutch schoolteacher called Hanny found a strange looking blue object that she had not seen before in classifications. She posted it on the forum and asked the astrophysicists what it was.

Voorwerp
The green-looking thing is the Voorwerp.
Credit for this picture goes to: NASA, ESA, W. Keel, Galaxy Zoo Team (Hubble Space Telescope).

Despite initially being sure that it would turn out to be nothing interesting, the astrophysicists investigated and found out that it was in fact something that they had never seen before. This object was called Hanny’s Voorwerp (voorwerp meaning ‘thing’) and after further analysis by Hubble, it is probably a cloud of gas and dust that had been heated up and excited by radiation from matter falling into a close galaxy’s black hole called a Quasar, which has itself now quietened down.

The GalaxyZoo community were enthused by this idea. Without prompting, they went through the images looking specifically for these voorwerpjes.

A similar thing happened with little green smooth galaxies, now called Green Pea Galaxies. They are unusual, as small smooth galaxies are normally red or orange, with older, cooler stars. They are now being investigated by astrophysicists.

GalaxyZoo is only one of the many projects available online where people have a chance to make their own scientific discoveries. PlanetHunters is a project designed specifically for this—analysing the light curves from stars and looking for the distinctive dip in the radiation that means a planet is blocking some of its light. So far they have identified 34 planetary candidates.

So, GalaxyZoo is now a community of people making discoveries and finding things out for themselves, clearly getting to know their galaxies well enough that they notice unusual features. However all you sceptics might point out that this seems like the enthusiasm of a few very keen people just educating themselves. The average user to the site can still go ahead and click away at the pretty pictures without learning anything.

Well, there is now a part of the website where you can open up all the images you have classified and perform your own analysis of them. There are plotting tools and tools for you to be able to analyse the colours of galaxies and find out more about them. You can actually do the science. This was all driven by the public by the way. Not by the scientists. Laypeople really are interested in this type of stuff.

There are also links on the site to scientific papers and a forum for discussion with experts. In his talk, Grant said that he wants to go further and write an online course that you can complete as you use the site to teach you about what you are actually classifying.

Anyway, even if people simply want to click at the pictures to help science, why shouldn’t they? Grant provided some interesting statistics in his talk about how we spend our collective time, for example that every hour 16 years of human effort goes into playing Angry Birds. Here is an even more crazy one: over 3 million years of human effort have gone into playing Call of Duty. Modern humans have only been alive for one fifteenth of that time.

Don’t worry! Obviously I am not saying that we should be playing less C.O.D… well maybe a bit less, but the point is that collectively we have a lot of time to waste, and if even a small fraction of that could be put into citizen science then the reward would be huge.

Not just for astrophysics too, which some of you may well regard as just as trivial and less fun than C.O.D—one of the things that surprised me about Grant’s talk was his excitement about the humanitarian projects that Zooniverse can provide. On the site now, for example there is a project run by Cancer Research where, by analysing pictures of breast cancer cell tissue, you can help spot cancer cells and see how they have responded to treatment. That is impressive. Instead of simply donating money to the charity, you can actually help with their research.

Zooniverse isn’t a company, so it survives on grants, but is always trying to expand. A new project being considered is the ability to upload satellite imagery of areas hit by disasters such as tsunamis and earthquakes within minutes of them occurring. Volunteers could then search through them online in a few hours and identify survivors and people in need of rescue and aid.

Crazy stuff!

The daily Zooniverse blog (which I think is run by Grant) can be found here:  http://daily.zooniverse.org/

Also thanks to Imperial Hub who organised the talk—this is the first one I have been to by them, but I hope to go to more.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.