The Study of Learner Topics in Reviews of Computer Science MOOC Courses

With the rapid growth of digital education, Massive Open Online Courses (MOOCs) have become an important tool for online learning due to their quality, accessibility, and openness. However, low participation and completion rates remain challenges in the MOOC landscape. This study focuses on analyzing user reviews of computer science MOOC Courses to identify common topics and sentiment patterns, providing insights for course improvement. A sample of 307 computer science courses from the Chinese University MOOC platform is used. The research utilizes text segmentation, embedding, dimensionality reduction, and clustering to extract thematic information from reviews, followed by sentiment classification. The analysis highlights that course content, learner experience, and instructors are the primary themes in the reviews, with content structure and the learning experience being major factors driving negative sentiments. Based on these findings, targeted recommendations for course optimization are provided.