Abstract:

The COVID-19 pandemic transformed the teaching style that had been implemented as the basis for face-to-face classes. Particularly in higher education institutions such as universities, remote lectures using teleconferencing tools such as ZOOM have become widely used as it has become difficult to conduct face-to-face classes in large classrooms. A feature of teleconferencing tools such as ZOOM is the ability to save remote lectures as video files such as the MPEG-4 format. These functions have facilitated real-time remote lectures using teleconferencing tools while simultaneously creating video materials for on-demand use. Against this backdrop, on-demand lectures using video distribution platforms such as YouTube are being widely implemented. Furthermore, efforts are being made to implement flipped classroom teaching using the created lecture videos.

When archives of remote lectures conducted in real time are used directly as teaching materials for flipped classroom teaching, it is difficult for learners to grasp which parts of the video content they should focus on. Real-time distance learning involves question-and-answer sessions and discussions between teachers and learners. Even if the frequency differs from that of face-to-face lectures, interaction occurs between the teacher and learners. This interaction between the teacher and the learner is useful for promoting the learner’s understanding in a class conducted in real time, regardless of the format, such as face-to-face or remote.

The flipped classroom is essentially a transfer of knowledge before class time. The knowledge transfer is performed in the class by having the learners watch the lecture videos in advance, and discuss what they have learned during class time. Therefore, the use of archives of real-time remote lectures as pre-study materials for flipped classroom teaching includes elements other than knowledge transfer, which may interfere with smooth knowledge transfer.

Video distribution platforms such as YouTube have the ability to analyze viewer-viewing trends. In addition to basic statistics such as the number of views, YouTube Analytics also offers functions which measure the engagement of viewers with video content. This includes the total play time and average watch time. For example, based on the average viewing time of the content, we can understand which parts of the lecture video learners are interested in. By understanding which parts of the video attracted the interest of more learners, we can gather information that will be useful for improving the video materials. By analyzing the results of the video distribution platform, we can extract pre-learning materials for flipped learning classes based on lecture videos, especially by improving the materials to increase engagement.

In this paper, we propose a method for analyzing lecture videos and improving video lecture materials by analyzing video distribution platforms. Based on the class archives of real-time remote lectures, we describe a policy for improving the class contents by applying the analysis function of the video distribution platform. We will also conduct an evaluation experiment to verify the effectiveness of the proposed method.

Keywords:

Flipped Classroom, Blended Learning, Course Material Amelioration, Course Material Analysis.