DiAna

DiAna(Dialog Analysis) 2016-

グループ学習における対話の分析を支援し,介入をウェアラブルデバイスを用いて行う情報システムの研究です.対面議論やオンライン議論の発話特徴量から,占有や交代の分析,内容の音声認識を行います.これらの成果をエージェントのファシリテーションによる参加支援,グループ学習への意外な観点の提供に活用します.

Diana(Dialog Analysis) is the research on an information system that supports the analysis of dialogue in group learning and uses wearable devices for intervention. We analyze occupancy and alternation from speech features of face-to-face and online discussions, and perform speech recognition of the content. These results are used to support participation through agent facilitation and to provide unexpected perspectives for group learning.

主要論文 Selected Papers

  • 西村龍之介,居原田梨佐,菅本祐也,石井裕,望月俊男,江木啓訓: 議論における発話の偏りに基づいて参加の均等化を促す議論支援システム.情報処理学会論文誌, Vol.65, No.1, pp.197-210, 2024年1月 [IPSJ]
  • Taisei Muraoka, Naruaki Ishikawa, Shigeto Ozawa and Hironori Egi: Effect of Presenting Co-occurrence Networks that Reflect the Activeness of Face-to-face Discussions, the 23rd International Conference on Human-Computer Interaction (HCI International 2021), LNCS12784, pp.347-360, 2021.7 (Online) [SpringerLink] [PDF]
  • 石川誠彬, 岡澤大志, 江木啓訓: 発話の占有を通知する議論訓練システムの提案, 情報処理学会論文誌, Vol.62, Vol.1, pp.64-77, 2021年1月 [IPSJ] [PDF]
  • Naruaki Ishikawa, Taishi Okazawa and Hironori Egi: DiAna-AD: Dialog Analysis for Adjusting Duration during Face-to-face Collaborative Discussion, The 25th International Conference on Collaboration Technologies and Social Computing(CollabTech2019), LNCS11677, pp.212-221, 2019.9 (Kyoto, Japan) [SpringerLink] [PDF]

TASS

TASS(Teaching Assistant Supporting System) 2016-

グループ学習における対話の分析を支援し,介入をウェアラブルデバイスを用いて行う情報システムの研究です.対面議論やオンライン議論の発話特徴量から,占有や交代の分析,内容の音声認識を行います.これらの成果をエージェントのファシリテーションによる参加支援,グループ学習への意外な観点の提供に活用します.

TASS(Teaching Assistant Supporting System) is the research on an information system to support the guidance of teaching assistants in programming learning. The system uses sensors to measure the position and posture of students while engaging in learning support activities in the classroom, and detects learners who need support and provides optimal support for each individual student. The system aims to change the behavior of teaching assistants through visualization and presentation of their characteristics.

主要論文 Selected Papers

  • Shin Ueno, Takahiro Yoshino and Hironori Egi: Addressing promotion system based on student data to supportesk-to-Desk Instruction by Teaching Assistants, the 13th International Conference on Learning Analytics and Knowledge (LAK23) Companion Proceedings, pp.162-164, 2023.3 (Poster)(Arlington, USA) [Web] [PDF]
  • 江木啓訓,横山裕紀,今村瑠一郎: プログラミング演習における学習支援方略に基づくTA支援システムの開発と実践, 情報処理学会論文誌 教育とコンピュータ(TCE), Vol.8, No.2, pp.1-11, 2022年6月 [IPSJ] [PDF]
  • 今村瑠一郎,照井佑季,上野真,江木啓訓: ティーチングアシスタントの学習支援状況を可視化するシステムの開発と実践, 日本教育工学会論文誌, Vol.46, No.1, pp.203-215, 2022年2月 [J-STAGE] [PDF]
  • Ryuichiro Imamura, Yuuki Terui, Shin Ueno and Hironori Egi: Introducing Real-Time Visualization Methods of Learning Support Behaviors for in-Classroom Lessons toward Optimized Assistance, the 11th International Conference on Learning Analytics and Knowledge (LAK21) Companion Proceedings, pp.133-135, 2021.4 (Poster) (Online) [Web] [PDF]

ThinkLeg

ThinkLeg 2018-

学習者が着席する机にセンサを設置して,脚部動作の計測により状況を推定する手法の研究です.オンライン学習や自習に取り組む学習者が,関心があるか,難しいと感じているか,精神的に疲労しているかなどを機械学習により推定します.結果はリアルタイムに可視化したり,効果的な休憩タイミングを提案するなどの方法で学習者にフィードバックします.

ThinkLeg is the research of a method for estimating the situation of a learner by installing sensors on the desk where the learner is seated and measuring leg movements. Machine learning can estimate whether a learner is interested in online learning or self-study, whether the learner is finding it difficult, and whether the learner is mentally fatigued. The results are visualized in real time, and feedback is provided to the learner in the form of suggestions for effective break times. (Both leg and reed are pronounced the same in Japanese "ashi"; a pun of "man is a thinking reed")

主要論文Selected Papers

  • Sei Yamamoto, Ryo Funabashi, Takeo Noda, Masataka Kaneko and Hironori Egi: An Approach of Multimodal Learning Analytics Based on the Distance between Learners' Heads in Collaborative Learning, the 14th International Conference on Learning Analytics and Knowledge (LAK24) Companion Proceedings, 2024.3 (Poster)(Kyoto, Japan)(Accepted)

CSCL

CSCL(Computer Supported Collaborative Learning) 2021-

協調学習における発話や身体動作を計測して可視化する,マルチモーダル分析の研究です.グループやペアでの学習活動における推移から,内面的状態を推定します.特に活動や理解の構造や転換点を情報システムで検出し,個別介入の必要がある状況を通知することによって学習支援に役立てます.

CSCL(Computer Supported Collaborative Learning) is the research of multimodal analysis that measures and visualizes speech and body movements in cooperative learning. We estimate the internal state from the transition in learning activities in groups or pairs. In particular, the structure and turning points of activities and understanding are detected by the information system and used to support learning by notifying situations that require individual intervention.

主要論文Selected Papers

  • Sei Yamamoto, Ryo Funabashi, Takeo Noda, Masataka Kaneko and Hironori Egi: An Approach of Multimodal Learning Analytics Based on the Distance between Learners' Heads in Collaborative Learning, the 14th International Conference on Learning Analytics and Knowledge (LAK24) Companion Proceedings, 2024.3 (Poster)(Kyoto, Japan)(Accepted)
  • Ryo Funabashi, Kohei Nabetani, Takeo Noda, Masataka Kaneko and Hironori Egi: Motivation Estimation Method for Computer Supported Collaborative Learning Using Tablet Computer, APSCE the 30th International Conference on Computers in Education (ICCE2022), pp.671-673(Vol.II), 2022.12 (Kuala Lumpur, Malaysia) [Web]
TOP