宮田 真宏

Considering that non-cognitive abilities develop significantly during childhood, the Japanese early childhood education community has recently begun to focus on the study of non-cognitive abilities. However, measurement and estimation methods for non-cognitive abilities face quantitative and objective challenges. In addition, there is little research on the active use of artificial intelligence technology in education. Thus, this study proposed a method for quantitatively estimating concentration in the classroom, one of the children's non-cognitive abilities. Specifically, it examined the number of intersections between children's gazes using information on the direction of their faces and their gazes during class and determined whether all children should focus their gaze on a small area. Moreover, it quantitatively examined whether each child behaved appropriately in different situations by measuring the angle between each child's gaze and the center of gravity of the intersections of all the children's gazes. As a result, we found that we could capture the characteristics of each child's behavior.

本研究では、教育方法の可視化に向けたICT開発に関する研究報告を行う。そこでは、子どもの顔の向き情報から推定される子どもの関心状態と授業デザイン（活動の流れ）の関連性ついて検討した。まず、小学1年生13名を対象とした授業の映像データ（6分18秒間）から、画像処理技術により子どもの顔の向き情報を連続的に抽出した。次に、得られた顔の向き情報から形成される交点群の空間的な広がりを円近似し、その半径の時間変化を求めた。最後に、その時間変化を授業デザイン（活動の流れ）及び教諭の働きかけと比較し、それらの関連性について検証した。その結果、交点の広がりの時間変化を表す波形に授業デザインの構造を反映する特徴が認められた。これにより、子どもの顔の向き情報により授業デザインの構造が定量的に可視化されることが示唆された。

In this study, we attempted to derive the "flow" and "rhythm" inherent in educational methods by using the intersection point group(IPG) data assuming the line of sight from the direction information of children in a group, as a trial for the development of technology to visualize educational methods. We applied a probability distribution modeling and Fast Fourier Transformation to IPG data that is annotated with an early childhood education video. As a result, a specific distribution pattern was observed in IPG as the three time intervals where characteristic changes in IPG distribution occurred. In addition, the similar frequency components were found in the distribution of IPG in the same time interval. When these results were compared with the qualitative evaluation of the instructional plan, it was confirmed that the "flow" of the activity design and the "rhythm" of the teacher's action in the educational method were related to the characteristics of IPG distribution. These results suggest that it is possible to visualize the "flow" and "rhythm" of educational methods from the IPG analysis.