Enhance E-Learning through Data Mining for Personalized Intervention

Abstract

E-Learning has become an integral part of college education. Due to the lack of face-to-face interactions in online courses, it is difficult to track student involvement and early detecting their performance decline via direct communications as we typically practice in a classroom setting. Hence there is a critical need to significantly improve the learning outcomes of online courses through advanced, non-traditional approaches. University courses are often conducted through a web learning management system, which captures large amount of course data, including students’ online footprints such as quiz scores, logged entries and frequency of log-ins. Patterns discerned from this data can greatly help instructors gain insights over students learning behaviours. This positioning paper argues potential approaches of using Data Mining and Machine Learning techniques to analyse students’ online footprints. Software tools could be created to profile students, identifying those with declining performance, and make corrective recommendations to instructors. This timely and personalized instructor intervention would ultimately improve students’ learning experience and enhance their learning outcome.