See Worlds No One Has Seen Before

Humans are far, far better at pattern recognition tasks than any computer. This is the basis of quite a novel effort in citizen science, GalaxyZoo. Astronomy has entered the era of vast datasets being produced by large-scale surveys such as the Sloan Digital Sky Survey, which in its first phase catalogued about 200 million celestial objects. Now astronomers now want to do studies that are equally vast in scope with these data. For instance, what would a sample of a million galaxies from SDSS tell us about galaxy formation and the nature of the early universe?

There is one slight problem with such studies — how many astronomers have time to look through a million images of galaxies? So they turned to the astronomy-loving public to help them with classifying galaxies. In mid-July, the GalaxyZoo site was opened. Participants are asked to tell whether an image shows a spiral galaxy, an elliptical galaxy, a pair of merging galaxies, or in fact the image is a star or some artifact. If the galaxy is a spiral, they are asked to tell if the direction is clockwise or anticlockwise.

The project has turned out to be phenomenally successful. By the first week in August, more than 80,000 participants had classified galaxies, the traffic being 20 times more than what the organizers were hoping for. And the astronomers are now assured of getting many replicate classifications for each image, always a good thing statistically.