Three-Party Interactive Tutoring System for Mastering Machine Learning

Interactive tutoring systems are expected to contribute to self-learning in various areas, such as machine learning. In this study, we develop an interactive tutoring system that performs tutoring in a three-party dialogue form using multiple virtual agents. This would induce lower psychological burden on the learner than one-to-one dialogues. To adapt this system to various levels of users, questionanswering functionality is essential. The proposed system generates answers to questions by using a knowledge graph automatically constructed from a textbook.