学習者成長モデルの構築とexMCRNNを用いた学習者データの分類手法の提案

カテゴリ: 論文誌(論文単位)

グループ名: 【C】電子・情報・システム部門

発行日: 2021/03/01

タイトル(英語): Construction of a Learner Growth Model and Learner Classification Method Using exMCRNNs

著者名: 林田　智弘（広島大学），木下　拓矢（広島大学），脇谷　伸（広島大学），山本　透（広島大学），西崎　一郎（広島大学），関崎　真也（広島大学），谷本　祐輔（広島大学）

著者名(英語): Tomohiro Hayashida (Hiroshima University), Takuya Kinoshita (Hiroshima University), Shin Wakitani (Hiroshima University), Toru Yamamoto (Hiroshima University), Ichiro Nishizaki (Hiroshima University), Shinya Sekizaki (Hiroshima University), Yusuke Tanimot

キーワード: WBT，一次遅れ＋むだ時間系，学習者成長曲線，分類，リカレントニューラルネットワーク　　Web-Based Training，first-order + deadtime system，learner growth curve，clustering，recurrent neural networks

要約(英語): In recent years, with the development of information technology, the use of Web-Based Training (WBT) and other individualized learning programs are increasingly used and are gaining attention. In individual online learning, efficient learning is possible by providing individualized learning materials based on the degree of understanding and growth characteristics of each learner. However, actually, appropriate teaching materials for each individual learner is not be provided, which reduces his/her motivation to learn and makes it difficult for him/her to learn. The reason is that it is difficut to establish a desirable relationship between an educator model on the learning support system and each learner. This paper proposes a learner classification method based on learners' growth curves using the neural networks for the purpose of providing appropriate learning support to each learner. Since a huge amount of data is required for training of the neural networks, this paper construct an “educator-learner” model based on a control engineering approach representing the interaction between learning support systems and each learner, and virtual learner data is generated. The usefulness of the proposed method is shown by numerical experiments.

本誌: 電気学会論文誌C（電子・情報・システム部門誌） Vol.141 No.3 （2021） 特集：スマートシステムと計測・制御技術 －超スマート社会に向けて－

本誌掲載ページ: 273-280 p

原稿種別: 論文／日本語

電子版へのリンク: https://www.jstage.jst.go.jp/article/ieejeiss/141/3/141_273/_article/-char/ja/