An educational model based on Knowledge Discovery in Databases (KDD) to predict learner's behavior using classification techniques

This paper examined the students' history of accessing the university Learning Management System (LMS) data. Classification techniques are used to build an educational model based on Knowledge Discovery in Databases (KDD) to predict learner's behavior. It identified the most valuable influencer for learning outcomes of the learners; it generated prediction models using the J48 decision tree algorithm and Multiple linear regression; and it determined how likely is a Distance Education (DE) learners to get a mark of “Passed” in a certain course which may offer vital information to the teachers and university administrators for program planning and learner support strategies. The proponents conducted experiments to predict the students' final rating based on their history of accessing the data in the university LMS. Based on the derived model, the score obtained from the participation in the online activities was the most valuable influencer for learning outcomes of the DE learners. Thus, the successful completion of the program depends on how the students interact with the activities posted in the LMS. As such, the generated model may be utilized to identify DE learners who need early intervention for better academic achievements and meaningful online learning environment.