Designing Learner-Centered Collaborative Learning by Incorporating AI-Based Teacher/Learner Agents with a Cognitive Model

Authors

  • Yugo HAYASHI College of Comprehensive Psychology, Ritsumeikan University Author
  • Shigen SHIMOJO Ritsumeikan Global Innovation Research Organization, Ritsumeikan University Author
  • Tatsuyuki KAWAMURA Ritsumeikan Global Innovation Research Organization, Ritsumeikan University Author

DOI:

https://doi.org/10.58459/icce.2024.4821

Abstract

This paper presents collaborative concept-mapping tutor (CoCot ver.2), a collaborative learning support system that integrates concept maps with a conversational agent. CoCot ver.2 features two agents: a teacher agent and a student agent. The teacher agent acts as a human instructor, engaging in conversations with learners, aiding their metacognition, and summarizing the discussion content. The student agent learns from the learners' concept map creation and generates their own concept map knowledge. These agents are developed using (1) a cognitive architecture (ACT-R) for knowledge generation for the agents' concept map and (2) GPT 3.5 for part of the language processing for agent-based feedback.

Downloads

Download data is not yet available.

Downloads

Published

2024-11-25

How to Cite

Designing Learner-Centered Collaborative Learning by Incorporating AI-Based Teacher/Learner Agents with a Cognitive Model. (2024). International Conference on Computers in Education. https://doi.org/10.58459/icce.2024.4821