Predicting student performance from their behavior in learning management systems

Abstract

Nowadays, Information and CommunicationTechnology (ICT) provides an opportunity to discover newknowledge and create a desirable learning environment. That iswhy the influence of ICT on education is irrefutable.Technology has changed the learning styles: the way peopleprefer to learn and improve the quality of their learning.Physical and online classes can be held concurrently so thatlecturers and students can interact via learning managementsystems. A Learning Management System (LMS) is anapplication software that plays a significant role in educationaltechnology. Such software can be designed to augment andfacilitate instructional activities including registration andmanagement of education courses, analyzing skill gaps,reporting, and delivery of electronic courses concurrently. Since all information and corresponding data are recorded andmonitored in the LMS, it can provide an accurate insight intostudent’s online behavior. In general, measuring studentperformance is an important part of the education system. Thefields of learning analytics and educational data mining bothemphasize the analysis of educational data in order to improveteaching and learning styles as well as to predict studentperformance. In the current study, we use data from theMoodle LMS from a collection of courses from a singleinstitution to identify weak/strong students during the course.The result has to be interpretable and understandable as theaim is to give this information to lecturers, who may use theinformation to improve their course and identify students whomay need special attention.