Toward crowd-sourcing social intelligent agents: Knowledge request-broker system and its implementation

A new approach to problem solving is introduced in this paper, featuring autonomous social agents, in which, simple agents with limited problem solving capabilities investigate a problem, digest small parts of it, and reunite in arbitrary social formations to represent the solution of the big problem as a whole. The social model is inspired by free formed business value-chain model, and agents behave on the basis of freedom and joy. The solution is implemented thanks to our knowledge request-broker architecture (KnoRBA), which facilitates better cooperation and mutual understanding between heterogeneous rational agents. Results show the successful construction of problem solving society by autonomous agents which efficiently solves a problem far beyond the capability of any single of them, which in turn confirms the usefulness of KnoRBA platform.