Modeling Group Creativity as the Evolution of Community-Level, Creative Problem Solving

Abstract

Understanding creativity and innovation in large communities (collectivities) is emerging as a key problem in computational and social sciences as the amount and importance of innovation achieved by groups significantly exceeds innovation produced by individuals. Community-level innovation raises new, intriguing yet challenging problems (e.g., knowledge and belief formation, learning, and decision-making at the community level). This chapter presents and contrasts computational modeling approaches for studying creativity in large communities, from cognitive models to multiagent models and machine learning models. It also details a new model for gaining insight on the evolution process that leads to innovative solutions by large communities. The model has the following components: (1) explicit knowledge representation, (2) expression of the evolution process, and (3) a backward differential reasoning method for characterizing the trajectory of the evolution process. The chapter presents an application of the proposed model as a cyber-social system for reducing fixation in communities working on related problems.