Cooperative learning on autonomous agents acquiring common language for action and perception

In order to perform shared tasks, all the participants in a multi-agent system (MAS) must agree common meanings for their perceptions and actions. Therefore, the basic communication and cooperation must be handled in a different way, mainly because the agents are autonomous components who sometimes make their own decisions. Keeping this in mind, a model for language acquisition and learning based on an underlying MAS is proposed. It takes into account issues concerned lexical acquisition and emergence capabilities as products of agents interacting in an artificial society. As a result, some simulations in which these agents are involved in a “talk” with others to agree common meanings are discussed