Effect of the Scaffolding Done by a BDI Agent on the Evolution of Learning Trajectories

Cognitive convergence between humans and agentive tools opens research opportunities to foster learning. The document presents an artificial cognitive architecture for an intelligent agent to perform educational scaffolding while the solution of the Towers of Hanoi game. The architecture’s design considers two paradigms: the theory of BDI agents (Beliefs, Desires, Intentions) and scaffolding pedagogical strategy. To validate the architecture, four game versions of the Hanoi Tower game were designed: game alone, game with graph, game with agent, game with agent and graph. Then, the game solution trajectories in the four environments were compared. 16 scholars aged 10 to 12 participated. The results suggest that the scaffolding carried out by the agent contributes to the evolution of schemes in the solution of well-structured problems.