Research on development of online teaching materials and supporting procedure for leamers to improve lecture teaching using information technology.

本研究的目的在于开发一套线上教材的制作程序，以及协助学习者改善课堂教学的辅助程序。首先，提出了利用自组织映射（SOM）对讲座视频的场景缩略图进行归纳的可行性。为了从一个讲座视频中提取一个场景缩略图的特征，对图像的15个片段进行离散余弦变换（DCT），然后从每个片段中提取3个主要分量。该系统被应用于一个序数讲座。其次，为了研究学生特征对学习绩效的影响，日本一所大学的46名新生和67名研究生完成了动机、个性和思维方式三个构念的调查，并参加了为期14周的两门混合学习课程，该课程由面对面授课和相应的在线模块和测试组成。记录学习成绩，如出席天数（NDA）、完成模块数量（NCM）和在线考试成绩（OTS）。运用统计分析揭示了特征、学习实践与学习绩效之间的关系。同样的调查也在美国一所大学进行，收集了常见的数据。第三，开发了在线学习者失败预测程序，用于在线课程的操作。该方法是基于不及格学习者的在线课程访问日志数据与已完成学习者的在线课程访问日志数据呈现出不同的趋势。此过程可应用于其他类数据。这些结果显示了利用信息技术改进大学教学并考虑学习者特点的可能性。

The purpose of this study is to develop a procedure for creating online teaching materials and supporting procedure for learners to improve lecture teaching.Firstly, we presents a feasibility that scene thumbnail images of lecture video can be summarized using Self-Organizing Map (SOM). To extract a feature of scene thumbnail images from a lecture video, the discrete cosine transformation (DCT) was conducted for 15 segments of a image, then 3 major components from each segment were extracted. The system was applied to an ordinal lecture.Secondly, to examine the impact of student characteristics on learning performance, 46 freshmen and 67 graduate students from a japanese university completed surveys on three constructs, which were motivation, personality, thinking styles, and participated in two blended learning courses which consisted of offace-to-face lectures with corresponding online modules and tests for 14 weeks. Learning performance as the number of days attended(NDA), the number of completed modules (NCM) and online test scores (OTS) were recorded. The relationship amongst characteristics, learning practice and learning performance are revealed using statistical analysis. Also, the same survey was conducted in a US university, common data were collected.Thirdly, a prediction procedure for failure of online learners was developed to operate online course. The procedure was based that access log data of the online course for failed learners showed different tendency with ones of finished learners. This procedure can be applied to other class data.These results show possibility of improving university teaching with information technology and consideration of learner's characteristics.