lördag 15 september 2012

Through the last couple of years, i've posted on various social networks about projects that blend gaming-like activities, or at least interactivity with real science projects. It's a subject that has grown together with crowdsourcing, and i find it greatly fascinating, that i can help scientists to new discoveries as a layman.
Most of these projects come down to machine learning, where you help identify positives and negatives in a data set to help the algorithm, or simply to guide scientists to interesting areas.
It has the same kind of appeal as the folding@ projects, where you lend out your cpu to science. But instead, here you lend out what makes humans so great, your mind and capability for pattern recognition.

It seems that a lot of the projects i've previously tried out are now gathered under one roof; https://www.zooniverse.org/

I've tried out a few of these projects, helping the Kepler telescope by looking for possible planets orbiting stars in Planet Hunters, and looking for green bubbles (possible star forming locations) for the Milky Way Project.

An interesting side note is in several of these projects, you can download the raw data for yourself, in case you would like to experiement with your own machine learning algorithms.