@techreport{oai:ipsj.ixsq.nii.ac.jp:00232904,
 author = {西川, 耕平 and 伊藤, 毅志 and Kohei, Nishikawa and Takeshi, Ito},
 issue = {16},
 month = {Mar},
 note = {ゲーム AI が人間のトッププレイヤを超え，プレイヤが自身の学習にゲーム AI を取り入れるようになっている．しかし，ゲーム AI を活用した学習支援において，最も単純な “AIの導いた答えを直接学習者に示す” という手法は，学習者から考える機会を奪ってしまい，却って学習効果が低下してしまうことが危惧される．本研究では，リアルタイムで進行するテレビゲームの「テトリス」を題材に，先行研究に倣った「未来局面の提示」という間接的な学習支援の手法を用いて学習を行うことで学習効果にどのような差異が生まれるかについて検証する．支援のレベルとして，ゲーム画面上にリアルタイムに AI が考える「1 手先」「3 手先」「5 手先」の未来局面を提示するシステムを作成した．支援レベルを調整した被験者群を用意して，学習させてそれぞれの学習効果を調べた．その結果，効果を期待していた「3 手先」「5 手先」という支援のレベルを抑えた間接的な学習支援において学習効果が見られず，「支援なし」群が最も学習効果が高くなった．その理由について，それぞれの被験者群の学習履歴をもとに考察を行った, Game AI has surpassed top human players, and players are now incorporating game AI into their own learning. However, it is feared that the simplest method of learning support using game AI, "directly showing the learner the answer derived by the AI," may deprive the learner of opportunities to think, thereby reducing the effectiveness of the learning process. In this study, using the real-time video game "Tetris" as a subject, we will examine the difference in learning effectiveness by using an indirect learning support method called "presentation of future position," which is in line with previous studies. As a level of support, we created a system in which the AI presents future game scenes of "next 1 move position," "position after 3 moves," and "position after 5 moves" in real time on the game screen. A group of subjects with adjusted levels of support was prepared and allowed to learn, and the learning effect of each was examined. The results showed that the indirect learning support with a reduced level of support, "position after 3 moves" and "position after 5 moves," which was expected to be effective, did not show any learning effect, and the "no support" group showed the highest learning effect. The reasons for this were discussed based on the learning history of each subject group.},
 title = {未来局面の提示によるテトリスの学習支援システムの提案},
 year = {2024}
}