Using Agents to Investigate Strategies for Human Collaborative

Given the cost and effort involved in investigating different methods or protocols for human collaborative learning, this paper proposes the use of software learning agents to simulate and test various protocols. The best performing protocols would then be tested on human subjects. The paper presents a new form of cooperative learning, called coactive learning. After arguing that humans could be taught to learn coactively, experiments are conducted with various coactive learning schemes. The experiments demonstrate situations in which machine learning agents using coactive learning can perform better than individual machine learning agents. Collaborative learning through coaction is shown to generate properties that individual learners can achieve only through computationally intensive strategies. It is suggested that human learners be trained to learn coactively using the most successful agent coacting protocols, and compared, on the same type of learning task, to individual learners and groups of learners not trained in these protocols.