Predicting Students’ Learning Outcomes in Massive Open Online Courses

E-learning has now become a common learning delivery method, especially in higher education during the Covid-19 pandemic. One of the obstacles faced by instructors through this method is the difficulty of observing the behavior of learning participants directly, even though the learning is presented synchronously, moreover, learning mode is presented asynchronously. However, currently, this problem can be coped with utilizing the field of Educational Data Mining (EDM). Some features or types of data sets are processed to produce important information from a learning management system. This type of dataset is divided into three categories, namely: 1) log file features 2) Forum Activities, 3) Evaluation activities feature. Predictive Learning Analytics (PLA) is one of the commonly used techniques in the field of learning analytics and EDM, this technique provides instructors with an understanding to predict the learning success of participants and how to find problems faced by them. This research focused on activities that can describe the students’ achievement.