Boosting Interactive Evolution Using Human Computation Markets

Interactive evolution, i.e. leveraging human input for selection in an evolutionary algorithm, is effective when an appropriate fitness function is hard to quantify yet solution quality is easily recognizable by humans. However, single-user applications of interactive evolution are limited by user fatigue: Humans become bored with monotonous evaluations. This paper explores the potential for bypassing such fatigue by directly purchasing human input from human computation markets. Experiments evolving aesthetic images show that purchased human input can be leveraged more economically when evolution is first seeded by optimizing a purely-computational aesthetic measure. Further experiments in the same domain validate a system feature, demonstrating how human computation can help guide interactive evolution system design. Finally, experiments in an image composition domain show the approach’s potential to make interactive evolution scalable even in tasks that are not inherently enjoyable. The conclusion is that human computation markets make it possible to apply a powerful form of selection pressure mechanically in evolutionary algorithms.