i202 Fall 2010 School of Information, UC Berkeley

In the age of information huge datasets aren't hard to come by. Everyone is trying to make sense of the rampant information made available to them. One of the largest (at least in terms of potential size) datasets to be analyzed is — somewhat ironically — the universe. It's big. 

The Galaxy Zoo project was launched back in 2007 with a goal of classifying a million galaxies using volunteers who would “determine by eye whether the galaxies are spiral or elliptical — a task for which computers are almost worthless” (Hand, 2010). The project takes advantage of the human brains' superior spatial intuition and awareness — something technology can't quite yet immitate. The site was hugely successful receiving more than 50 million classifications in it's first year alone. With hundreds of thousands of users, the data provided was enough to prove that the project's classifications “are as good as those completed by professional astronomers.” (galaxyzoo.org)

spiral galaxy (left), elliptical galaxy (right)

Based on the success of the initial project, the simple classification scheme was expanded for a second  project, Galaxy Zoo 2, focused on gathering more detail about a subset of galaxies. More recently, the project is expanding to take a deeper look at the universe. The first two projects focused on galaxies that were (relatively) nearby. This year, for its 20th anniversary, Hubble released 200,000 images for Galaxy Zoo users to classify (Cooney, 2010). These galaxies are much farther away, essentially allowing us to look back in time (since the light they emitted has taken millions or billions of years to reach us). Effectively the new Galaxy Zoo Hubble project is collecting data on another dimension — time — allowing astronomers to build a better history of the evolution of the universe. Awesome.

Galaxy M100 (Hubble)