Human Computation Games and Optimization of Their Productivity

Web 2.0 technology enables people worldwide to collaborate over the Internet, a phenomenon known as social collaboration. While the incentives for social collaboration are primarily enthusiasm for a particular subject, building a reputation, or gaining a benefit by doing something in exchange for using services or downloading files, the emergence of human computation games has shown that the prospect of having fun can be a strong incentive for participants to actively engage in such collaboration. Among the human computation games, ESP game (ESP stands for Extrasensory Perception) is one of the most popular ones. To play an ESP game, two randomly matched players assign labels that appropriately describe an image provided by the system. It has been shown that the "outcomes" of ESP games have many useful applications, such as image-based CAPTCHA tests and semantic image searches. In this chapter, we provide an overview of human computation games and present an analytical model for computing the utility of ESP games, i.e., the throughput rate of appropriate labels for given images. The model targets generalized games, where the number of players, the consensus threshold, and the stopping condition are variable. Via extensive simulations, we show that our model can accurately predict the stopping condition that will yield the optimal utility of an ESP game under a specific game setting. A service provider can therefore utilize the model to ensure that the hosted ESP games produce high-quality labels efficiently, given that the number of players willing to invest time and effort in the game is limited.