Automatic Classification of Instructional Video Based on Different Presentation Forms

Abstract

The number of instructional videos is increasing rapidly in the digital age, and the presentation forms of the videos are different. To allow learners to select suitable instructional videos more quickly and improve learning efficiency, the automatic classification of instructional videos becomes significantly important. This paper presents an automated instructional video classification method based on Yolov4 target detection network model and Naive Bayes classification algorithm. Classification rules are determined according to instructional videos presented in different forms, and then key frames are extracted based on inter-frame difference. Finally, instructional videos are classified according to key frames of videos. The experimental results show that our automatic classification method for instructional videos based on different presentation forms can achieve an accuracy of 95%, which is helpful to promote individual learning and optimize online learning experiences.