Hybrid Analysis of the Learner's Online Behavior Based on Learning Style

Since the COVID pandemic, universities propose online education to ensure learning continuity. However, the insufficient preparation led to a major drop in the learner's performance and his/her dissatisfaction with the learning experience. This may be due to several reasons, including the insensitivity of the virtual learning environment to the learner's preferences. We propose to address the issue of student's dissatisfaction and lack of interaction, by integrating learning style theory into the analysis of the learner's online behavior. Our work differentiates itself from the rest of researches that employed learning style theory by its two step process. First, we classify the learning activities into learning categories based on learning style theory. Second, we define behavioral features that quantify the learner's behavior across the learning categories. The analysis of the learner's online behavior based on the behavioral features revealed new aspects of the learner's preferences. We consider these findings to be best useful for developing learning style-sensitive adaptive learning environments. Nevertheless, the behavioral features could be beneficial in different contexts. In fact, when applied to course outcome prediction, the behavioral features enhanced the results by 10%. The latter indicates that behavioral features reflected the correlation between behavior and academic performance.