Intelligent Educational Video Recommendation System based on Felder Silverman Learning Styles Model

Abstract

ABSTRACT This thesis develops and evaluates an automated system for classifying software engineering educational videos based on the Visual/Verbal dimension of the Felder-Silverman Learning Style Model. As online educational content proliferates, matching videos to individual learning preferences becomes crucial, yet manual classification is impractical at scale. Following a Constructive Research Approach, this study addresses this challenge through computational detection of learning style characteristics in videos. The system extracts visual features (diagram and animation density) and verbal features (speech ratio and text density) using computer vision and audio processing techniques. Evaluation with 20 software engineering videos shows good classification accuracy compared to expert assessments, with visual features contributing approximately 65% to classification decisions and verbal features 35%. The system's confidence scores correlate positively with classification accuracy, enabling identification of potentially unreliable classifications. Further evaluation with 40 undergraduate students (20 in a test group using the personalized system and 20 in a control group using non-personalized recommendations) examines the effectiveness of personalized video recommendations and system usability. This research contributes to the field of video-based learning by demonstrating the viability of automated learning style classification for educational videos. Furthermore it contributed to the field of automatic video analysis by developing an efficient sampling methodology, identifying key classification features, and creating a practical system to support personalized learning in software engineering education. Future work could extend this approach to other learning style dimensions and incorporate machine learning techniques to further improve classification accuracy. Keywords: Learning Style Classification, Felder-Silverman Learning Style Model, Educational Video Analysis, Personalized Learning, Computer Vision, Feature-Based Classification