A Multiagent Theory of Agency and Learning based on Imitation

G. Trajkovski (USA) and G. Stojanov (Macedonia)


Robotics and Automation, Cognitive Robotics, Developmental Robotics, Multiagent Systems,Cognition, Perception.


This paper presents the multiagent extension of our uniagent theory of agency, based on the concepts of interaction with the environment and expectation. Motivated by studies in animals, the interaction between the agents is carried out via imitation. The drive to imitate is the top of the drive hierarchy of the agent. During the imitation convention, they exchange their contingency tables. Several social consequences from the perspective of learning and emergence of language are discussed. The paper also overviews the preliminary results of the ongoing project in this area called POPSICLE (Patterns in Orientation: Pattern-Aided Simulated Interaction Context Learning Experiment), designed to illustrate the major aspects of the MASIVE (Multi-Agent Systems Interactive Virtual Environment) theory.

Important Links:

Go Back