Learner Level and Preference Prediction of E-learners for E-learning Recommender Systems

Abstract

An effective e-learning system must identify learning content appropriate for the needs of the specific learner from among the many sources of learning content available. The recommendation system discussed here is a tool to address such competence. Identifying learner levels, and thereby identifying the appropriate learning content, is possible only if the learning content is prepared using a proven instructional strategy which covers various learner levels. Therefore, in the proposed method, the learning content is prepared using David Merrill’s First Principles of Instruction, a problem-based approach that has four phases of instruction: activation, demonstration, application and integration. These four phases are used to predict learner levels for three different types of media content, namely text, video and audio, and to determine a rating. The naïve Bayes classifier assumes that the presence of a particular feature in a class is unrelated to the presence of any other feature. Learner ratings are used to predict learner preferences as to the type of content. To identify the learner’s level, the learner rating and the instructional phases which they prefer most are used. The same classifier is used to identify the level and preferences of the learner. To estimate predictive accuracy, a k-fold cross-validation technique is used. The experimental results show that the proposed classifier yields maximum accuracy of 0.9794 and a maximum kappa statistic of 0.9687 for learning level and preference, respectively.