Home // International Journal On Advances in Intelligent Systems, volume 14, numbers 1 and 2, 2021 // View article
Authors:
Yoshimasa Ohmoto
Hanako Sonobe
Toyoaki Nishida
Keywords: Human-agent interaction; metacognitive suggestion; insight problem solving.
Abstract:
For an agent to teach a person a problem-solving attitude by giving him advice that does not directly contribute to solving the problem, a strategy that considers changes in the person's long-term attitude must be designed. This study aimed to investigate how the mental state of participants performing a task is affected during short-term and relatively long-term periods when they are advised either based on their conditions or mechanically at regular intervals. We focused on metacognitive suggestions during insight problem-solving as an example of advice that would be effective even if given by the agent. By these means, the effect on the human mental state over a relatively long period of time when the agent gives advice is examined. We conducted an experiment using two types of suggestion agents and observed that participants were likely to accept metacognitive suggestions provided by an agent when the suggestions were given based on an inner-state estimation of the participant. An analysis of mental state changes based on physiological indices suggested that the use of metacognitive suggestions by agents based on participants' conditions affected the mental state in problem-solving activities in the short and long term. It is also suggested that if the advice is not given depending on the situation, the effect of the advice in mitigating the impasse reduces as the task progresses. These findings will contribute towards the implementation of a tutoring agent.
Pages: 36 to 45
Copyright: Copyright (c) to authors, 2021. Used with permission.
Publication date: December 31, 2021
Published in: journal
ISSN: 1942-2679