Using Machine Learning to Predict the Low Grade Risk for Students based on Log File in Moodle Learning Management System

Abstract

The ever-growing demand for online teaching and studying requires the deployment of an effective online learning model. With diminishing face to face learning opportunities, schools are now heavily invested in improving student's sense of active learning and ultimately reduce the rate of student failures and dropouts. Currently, researchers have aimed to build a solution to analyze learners' behavior from the data collected through the online learning site - Moodle LMS and use the Linear Regression algorithm to predict the learning outcomes upon the completion of a student's course. The purpose of the study is to provide lecturers with a set criterion to standardize student learning outcomes during in the teaching process. On that basis, the lecturer can filter out the list of students who are at risk of failing the subject, and promptly warn students to change their learning attitude more actively, so that students can achieve satisfactory results. at the end of the course, thereby reducing the rate of students failing and dropping out of school.