Massive open online courses (MOOCs) have often been criticized for providing limited opportunities for active learner participation, resulting in low levels of social presence and engagement. To address this issue, this study aimed to develop a learne...
Massive open online courses (MOOCs) have often been criticized for providing limited opportunities for active learner participation, resulting in low levels of social presence and engagement. To address this issue, this study aimed to develop a learner-participatory MOOC using AI edutech. Learning analytics were employed to diagnose problems in an existing course and to derive design strategies for the subsequent development process. This study adopted a design and development research methodology based on the RPISD model, involving analysis, design, prototype development, and two rounds of usability testing. As a result, a course syllabus, weekly lesson plans, an AI-based quiz tool, and a public chat space equipped with an AI chatbot were developed. The learning analytics results confirmed several issues in the course offered in the previous semester. Specifically, learning frequency declined over time, question-and-answer and discussion activities were absent, and several learners completed the course by attempting only the quizzes without viewing the instructional videos. To address these issues, a chat space equipped with an AI chatbot was developed to facilitate question-and-answer and discussion activities and to provide course-related guidance. In addition, an AI-based quiz tool was developed to enable repeated attempts on quiz items, thereby promoting mastery learning. This study presents an empirical case of developing a learner-participatory MOOC using AI-based educational technology grounded in an objective diagnosis of the course through learning analytics.