Analyzing and Scaffolding Semi-Structured MOOC Forums to Support Social Learning and Help-Seeking

• Main Author: Xu, Yiqiao
• Format: Dissertation
• Language: English
• Published: ProQuest Dissertations & Theses 01.01.2022
• Subjects: Data mining; Distance learning; Educational technology; Intermediate Grades; Learning Theories; Literature Reviews; Online instruction
• ISBN: 9798358417618

Massive Open Online Courses (MOOCs) are purely online learning environments that are typically available for free for anyone with internet access. These courses have attracted hundreds of thousands, if not millions of learners around the world and the number of users has increased rapidly along with the growth of new platforms and a wide diversity of courses. MOOC classrooms are typically structured around a common set of educational tools including a basic learning management system, video lectures, quizzes, and a discussion forum. In MOOC, discussion forum is the only medium for student communication, which records a rich amount of massive data on students’ activities. Via students forum interaction, we can understand what their learning behaviours are and how these learning behaviours impact their course performance. Understanding students’ behaviours can also help instructors provide useful interventions and improve students’ learning gain.In this work, we analyze students’ discussion forum activities in several offerings of one MOOC which was delivered on two different platforms. We collected students interactions from the course platforms and the discussion forum to explore their social communication and learning behaviours. To organize the semi-structured forum, We first examine how MOOC students build their social community. Then we model students’ learning behaviours based on their discussion forum post content. Eventually, we automatic generate interventions for instructor and suggestions for students based on analyzing students post content.Organizing semi-structured MOOC discussions can help us to improve students’ educational experiences in many ways. First, by analyzing student weekly online activities, we track evolution of study over time by topic modeling algorithm. Second, via language analysis on posts content, we provide instructors with a dashboard that shows the evolution of topic over time. Last, we respond to posts (particularly questions) with pointers to course materials and current course posts that are relevant. The results of our work can provide instructors with insights into students online behaviours from the semi-structured discussion forum and also provide students supports to improve their learning gain.