Rough set and machine learning approach for identifying flow experience in e-learning

Abstract

Flow experience as a psychological theory has been implemented in various fields, especially in games and marketing. Flow is an optimal condition that someone feels immersed, focused and enjoy an activity. Based on this theory the psychological experience is represented as anxiety, boredom, and flow. Considering the psychological conditions in learning activities can improve the performance of students. Thus, involving the flow experience in learning process particularly in e-learning environment becomes very interesting and challenging. This challenge is how to identify the student flow experience during interaction students with the e-learning. In the previous study, flow was measured by conducting a series of questionnaires after a learning process. This is very inefficient, especially if applied in an adaptive e-learning. Additionally, this measurement is often unnatural, since it cannot capture the students learning behavior. Therefore, this study presented flow experience identification when students interact with e-learning using rough set and machine learning approach. Rough set is an efficient tool to solve the uncertainty, imprecision, and vagueness. While, the fuzzy rule and decision tree, as part of machine learning methods were implemented as a comparation. The identification was done using learning behavior parameters included duration of access, frequency of access, assessment score, and duration time to complete the assessment. As the results, the rough set can identify the flow experience with accuracy level is 92.92%. On the other hand, the fuzzy rule and decision tree provide accuracy level are 91.86% and 92.39%. As a conclusion, this study showed that the flow experience could be measured with the high accuracy level. In the context of e-learning, it can be used by e-learning to provide an adaptation. Appropriate adaptation is expected can keep the psychological condition of the students in optimal state.