Crowdsourcing Science

This article first appeared in our Michaelmas Term ‘Innovation Issue’, but has since been amended following a request from Professor Chris Lintott, a researcher at the University of Oxford’s Astrophysics Department. We would like to take this opportunity to apologise to Professor Lintott for mis-representing the quote he gave for the original article. This version more accurately highlights his thoughts on the matter of science crowdsourcing and we would like to thank him again for his participation in the interview.

A cure for cancer, the discovery of intelligent life elsewhere in the Universe, and an encyclopaedic understanding of the human brain may seem like science fiction fodder, but they may be closer than we think. The answer, according to many scientists, lies in plumbing “big data”. In this issue, we talk to Chris Lintott, Professor of Astrophysics at Oxford University and Co-Presenter of the BBC’s The Sky at Night, about his experiences as a leader of the Zooniverse project, which recruits members of the public with no formal scientific training to the task of mapping distant galaxies.

Professor Lintott’s first immersion in the field of big data came nearly a decade ago, when he arrived in Oxford as a researcher. In 2006, he and his colleagues quickly realised that they wanted to classify images of almost a million galaxies captured in the Sloan Digital Sky Survey launched at the turn of the century. This project had amassed more data in its first few weeks than all that previously collected in the history of astronomy; although the use of supercomputers made data collection simple, a lack of manpower was holding up efforts to accurately map the skies above us. Professor Lintott devised an innovative answer to this problem – crowd-sourcing the effort and recruiting the necessary manpower online. He created Galaxy Zoo, which allowed ordinary people to sign up and classify galaxies themselves. People with no relevant knowledge could sign up and identify galaxies in their spare time with just some basic training and the click of a mouse. Galaxy Zoo is now part of Zooniverse, a collection of citizen science projects ranging in focus from papyrology to pathology. The scheme has over a million volunteers, or ‘zooites’, signed up worldwide. Some exciting discoveries have already been made, including that of a planet with four suns, and the prospects for this approach seem highly favourable.

So what’s next? Crowdsourced science has already come a long way. In the early days, it was a lack of effective computer processing power that represented the major hurdle in astronomy research. The launch of SETI@home in 1999 went some way to addressing this problem, combining the donated power of computers across the world to create a supercomputer capable of analysing radio signals in the search for signs of intelligent life. A year later, Folding@home was created, with the aim of characterising the 3-D structures of proteins. Misfolded proteins have been linked to various neurodegenerative diseases, and this project hopes to advance research in this field, and also in oncology and basic drug design. Since its launch, Folding@home has contributed to the production of 114 scientific papers. While these projects can engender a sense of involvement among their members, the passive nature of their contributions often leaves some craving more. Zooniverse soon gave this to the public, allowing people to become more actively immersed in the science that lay behind the research. This has since been built upon by the ‘foldit’ project, launched in 2008 by Washington University. Members of the public can download software designed to mimic a computer game. The aim is to find the most likely formation of an enigmatic protein, refractory to conventional techniques such as X-ray crystallography, uploaded by a stumped research team. This innovative solution has already achieved significant success – in 2011, the structure of an AIDS-related enzyme was solved after 15 years of unsuccessful graft from various research groups.

Crowdsourced projects have spent the past decade becoming more and more complex and yet they are likely to start becoming simpler. The surge of smartphones has created a vast pool of untapped processing power. The HTC app ‘Power to Give’ and the Samsung equivalent ‘Power Sleep’ have both been released recently to put idle CPUs to use, working in the same way as SETI@home and Folding@home. In addition, many of us play games on our phone already – so we might as well be contributing to research while doing so. In response to this, Cancer Research UK released the world’s first free mobile game that contributes towards oncology research.

So why do people join crowdsourcing efforts? “Fame and glory?”, jokes Professor Lintott. “Most of the studies we’ve done…tell us that people want to take part because they genuinely get a kick out of spending a few minutes to do something useful, to make a contribution to science or to research. I think that’s fabulous! This idea – that people want to do something useful, that they want their contributions to count – is at the core of why me and my team spend so much time on Zooniverse, and I find it inspiring”.

With big data now an area of special scientific interest, particularly in the context of cancer biology and the connectome (an exhaustive description of the trillions of collections formed by neurons in the human brain), crowdsourcing could hold the answer to enigmatic questions that have plagued scientists for decades. Perhaps one day, the next big breakthrough will be made by a student idly playing a game on the bus to school.