A Domain-Driven Framework to Analyze Learning Dynamics in MOOCs through Event Abstraction

Interest in studying Massive Online Open Courses (MOOC) learners’ sessions has grown as a result of the retention and completion issues that these courses present. Applying process mining to study this phenomenon is difficult due freedom navigation give their students. The goal research provide domain-driven top-down method enables educators who are unfamiliar with data analytics search for set preset high-level concepts own MOOC data, hence simplifying use typical techniques. This accomplished by defining three-stage generates low-level event log from minimum model then abstracts it seven possible learning dynamics student may perform session. By examining actions students successfully completed Coursera introductory programming course, framework was tested. As consequence, patterns repetition content assessments were described; discovered students’ willingness evaluate themselves increases they advance through course; four distinct session types characterized via clustering. shows potential employing abstraction strategies gain relevant insights educational data.