@techreport{oai:ipsj.ixsq.nii.ac.jp:00223782,
 author = {森田, 孝裕 and 張, 亮 and 長手, 厚史 and Maryam, Alimardani and 西尾, 修一 and Takahiro, Morita and Liang, Zhang and Atsushi, Nagate and Maryam, Alimardani and Shuichi, Nishio},
 issue = {25},
 month = {Jan},
 note = {「トピックの難易度」と「説明の上手さ」が異なる 4 種のプレゼンテーション動画視聴時の脳活動を脳波でセンシングし，脳波データから複数パターンの特徴量を生成後，2 種別間の差分評価を行った．6 件の評価ケースで検定を行った結果， 2 種別間を区別しやすい特徴量が評価ケース毎で異なることを確認した．また，機械学習を用いた正答率評価では，実験参加者 16 名に跨った識別器を生成したところ，最大で 71%（チャンスレベルから +21%）の正答率を記録し，更に，θ波・α波・β波から生成した複数特徴量を同時利用した場合の優位性も確認した．, EEG was used to sense brain activity when viewing four types of presentation videos with different "topic difficulty" and "explanation skills", and after generating multiple patterns of feature values from the EEG data, the difference between the two types was evaluated. As a result of Wilcoxon test with six evaluation cases, it was confirmed that the feature value that makes it easy to distinguish between the two types differs for each evaluation case. In addition, in the evaluation of the correct answer rate using machine learning, when we generated classifiers for 16 experiment participants, we recorded a maximum correct answer rate of 71% (+21% from the chance level), and we also confirmed the superiority of using multiple feature values generated from θ, α, and β waves at the same time.},
 title = {特徴が異なるプレゼンテーション動画視聴時の脳活動比較－トピックの難易度と説明の上手さを指標とした評価－},
 year = {2023}
}