The growing network of people, through telecommunications, smartphones and the Internet allows people to come together in many ways. People share aspects of their lives in numerous ways and access unprecedented amounts of information, content and services. In recent years, however, it has become apparent that this vast network of connected and on-line people are not just consumers, but a valuable resource. ‘Crowdsourcing’ is the term used for outsourcing tasks to large groups of people, and today’s massively connected world makes crowdsourcing more viable and powerful than ever before.
Recently, we have seen some fun and exciting applications of crowd-sourcing in scientific research. Classification tasks, or repetitive tasks that cannot be performed by a machine are particularly suitable for crowd-sourcing. The idea is that certain tasks that may usually be performed by teams of people or by computers can be broken up into many parts, so everyday people can offer a few minutes of their time for free.
MoonZoo is such a project (read an article on Wired), focusing on classifying features of the universe, such as the formation of galaxies and the surface of the moon. The project aims to classify features in the gigantic archive of photos taken by astronomy satellites, thus extracting invaluable information. Users of the site look at photos of the moon surface and identify and locate craters, boulders and other features by comparing them with examples provided. GalaxyZoo uses the same approach to classify types of galaxies – human input is invaluable, elegantly put ”Computers will slowly get better at classifying galaxies, but looking at an image and asking ‘what’s that odd thing?’ remains uniquely human.” - Galaxy Zoo. Another interactive science crowd-sourcing project, Foldit, allows users to solve visual puzzles folding proteins to help discover ways of making chemical reactions more efficient. For more examples of science crowdsourcing projects that you can try look here.
A more common example that you are likely to have come across is “captcha” the security check on many websites that asks you to spell out difficult to read words. A basic nuisance to most, this process is actually helping the process of digitising books – words that cannot be correctly identified by the scanning software are crowdsourced by getting people to identify what they are.
But crowdsourcing has many uses beyond science and academia. Mobile apps are emerging that can take advantage of their users, to help eachother, or to help businesses. MapShare, Trapster, SubwayArrival and GoogleMaps all allow users to share information about transport and traffic to promote safety on the road or help others avoid congested areas. LocalMind allows people to share information about what’s going on around them – such as whether a particular bar is busy. Crowdsourcing is a great way to quickly acquire up-to-date information about an area or situation, and has recently been employed in many humanitarian and natural disasters.
Businesses already value the extensive customer behaviour tracking that is now possible because of our extensive digital lives. Facebook has become a useful source of information for businesses to perform consumer research, with many people wilfully publishing information about their behaviours and preferences that they would previously have had to go to great lengths to get via customer surveys.
Here in the Scribble offices, we’ve been playing the massively popular ‘DrawSomething’ app – a social game similar to Pictionary. With millions of people drawing their interpretations of specific words, could this be an interesting resource: for example to analyse social trends, or to help train computers to recognize images? Game developers have recently started to monitor detailed stats about game-player’s habits to inform design changes.
Crowdsourcing offers businesses and app developers a new way of looking at how they relate to their customers – users can be viewed as a resource and not just consumers. Maybe the way people use their apps can offer interesting insights – so what exciting uses will we see for crowdsourcing apps in the future?