target="_blank" rel="nofollow" href="http://www.tetris.com/">; last retrieved May 2014.
2. http://www.rubiks.com/; last retrieved May 2014.
3. http://brainage.com/; last retrieved May 2014.
4. http://bigbrainacademy.com/; last retrieved May 2014.
2 Related Literature
This book is related to several bodies of existing literature, including volunteer computing, human computation, serious games, computational biochemistry, and visualization and interaction.
2.1 Volunteer Computing and Human Computation
Volunteer computing is a method by which volunteers are able to donate their computer’s spare time and space to various projects. The volunteer computing model has risen in popularity recently, and has allowed scientists access to unprecedented amounts of computational power. One of the oldest and largest volunteer computing projects is SETI@home1 [Sullivan III et al. 1997 ]. This project uses a screensaver to analyze radio telescope data. There is an open source platform for developing volunteer computing projects, the Berkeley Open Infrastructure for Network Computing (BOINC),2 which allows users to manage and share their computer’s resources between the many projects using the platform [Anderson 2004 ]. BOINCprojects have a variety of goals, from climate prediction [Stainforth et al. 2005 ] to searching for pulsars [Knispel et al. 2010 ].
By using the volunteer computing model, projects not only gain access to massive computation, but also allows the public to make contributions to science. However, with this model, their contributions are mostly passive—they don’t even have to be at their computer. This work aims to use not only the power of networks of computers, but also that of networks of humans, and allow people to make active contributions to science.
There has been work recently on leveraging a human workforce for computational tasks that computers are not yet able to perform satisfactorily. A more general field of “human computation” or “distributed thinking” is emerging. On a smaller scale, augmenting automated heuristics with interactive human input can help to solve basic spatial problems [Anderson et al. 2000, Lesh et al. 2005 ]. On a larger scale, general tasks desired for humans to perform are posted online, and users can determine which tasks they would like to perform. Amazon’s Mechanical Turk3 is one example of such a system, where users are actually paid to perform tasks. Example tasks include translation of text, rating search results, and determining the tone of an article. Bossa is an open source system for managing similar user tasks.4
In this context, there has been much interest in using games as a means of motivating people to perform tasks that are currently difficult for computers. One particularly active area is in computer vision and image recognition. Humansare particularly adept at reading words in images and determining the objects in a scene, when compared with current computational methods. The difference in ability is strong enough that vision-based tests are often used as a proof of humanity with CAPTCHAs [Ahn et al. 2003 ].
Games such as the ESP game [von Ahn and Dabbish 2004 ], Peekaboom [von Ahn et al. 2006 ], and Google Image Labeler5 use human image-recognition ability to produce labeled images from gameplay. Image recognition has also been used for finding particular features of interest in scientific data, such as looking for signs of interstellar dust [Westphal et al. 2010 ], measuring and aligning features on a planet’s surface,6 and classifying galaxy shapes.7 These projects have been successful in motivating players to sift through large image sets, which would otherwise be a mundane task.
Some games have approached other types of problems. Pebble It8 is a game which studies human solutions to the graph pebbling problem, with the goal of developing better algorithms to solve it [Cusack et al. 2006 ]. Outside of games, some work has examined how to fit human problem solving into various optimization problems [Anderson et al. 2000, Lesh et al. 2005 ]. This work is different because it leverages a deeper human problem solving ability to create interesting scientific results.
2.2 Serious Games and Gamification
Recently, a field known as “serious games” has been identified. The most general definition is any game that has a purpose beyond simply entertaining the player; however, it often connotes games whose purpose is training or education. The line between game and simulation or application is also not always well defined. A game-based approach is appealing because games are meant to be engaging and motivating. Furthermore, other fields are taking advantage of the fact that gaming has pushed the limits of interactive simulation and authoring technology, such as 3Dengines [Susi et al. 2007 ]. Some of the wide-ranging applications of serious games are firefighter training [Backlund et al. 2007 ], raising awareness of social issues,9 and military recruitment.10 In this book, the main goal is to generate useful scientific discoveries; however, other aspects of game design, such as the requirement that the game be fun, contribute to achieving this goal, as the results rely on players playing the game.
One particular subgenre of serious games is games for health. Playing games has, in some settings, been shown to be beneficial to the player’s health. Games have been shown to be useful for rehabilitation, development, and therapy, and even for distracting patients from pain or bad habits [Adriaenssens et al. 1988, Griffiths 2005 ]. Nintendo’s Wii Fit package11 is intended to help players improve their personal fitness.
Many games emphasize social interactions as well. Massively multiplayer online games (MMOs), like World of Warcraft12 and Second Life,13 often host persistent virtual worlds where players can customize avatars, socialize, and work together with other players. Diverse niche MMOs exist, targeting teens or people interested in racing, allowing people with similar interests to interact [Zenke 2008 ].
Similarly, Alternate Reality Games (ARGs) engage large groups of people to participating in narratives in the real world [Martin et al. 2006 ]. Often, multiple forms of technology will be used to coordinate the players, who will be working together towards a common goal. I Love Bees, a popular ARG, had players work together to find payphones and answer prerecorded questions [Terdiman 2004 ].
2.3 Computational Biochemistry
In the field of biochemistry, many computational methods have been used to study protein folding, predict protein structures, and design new proteins. The most closely related to our work is Rosetta. Rosetta combines structural energy minimization with a Monte Carlo search algorithm to predict native protein structures [Rohl et al. 2004 ]. Foldit uses the Rosetta software for its energy function, as well as many of the algorithms and functionality for energy minimization and protein manipulation. In order to access massive amounts of computation for searching a protein’s large structural space, the volunteer computing project Rosetta@home14 runs Rosetta’s algorithms on volunteer’s computers.
Another volunteer computing project, Folding@home,15 aims to simulate the process of protein folding. The Folding@home project has also been ported to the PS3, and thus has access to powerful hardware and gives gamers the opportunity to help science. Folding@home’s approach is based on simulating the molecular dynamics of protein folding [Pande et al. 2003 ]. Onedifference to note is that Rosetta and Foldit only attempt to determine the final structures, while Folding@home simulates the folding process. This makes Rosetta and Foldit more amenable to problems like protein design, where one is interested in the final folded structure, while Folding@home’s approach is more useful for studying topics like protein misfolding.
Many visualizations for biological molecules have been developed to aid biochemists. Popular visualizations include Corey-Pauling-Koltun