Students have different learning styles when studying online. Meanwhile, lecturers use the same method for all students who take their online lectures. These different learning styles can affect the level of understanding and the results obtained by students. By knowing student learning styles, lecturers are expected to be able to use the right way in delivering material. In this research, we developed a student behavior analysis feature on self-developed Virtual Learning Environment (VLE) called Enterprise Hybrid Online Learning (ETHOL). Students’ data collected includes data on online activities, personal data, and survey data on student learning styles. User behavior analysis was carried out by dividing into three clusters: average scores, time to collect assignments, and student learning styles. The clustering method used is the Hierarchical K-Means. The results obtained are students who have the habit of collecting assignments on time have higher scores than others. In addition, the lecturer is able to see the results of the analysis of the behavior and learning styles of each student. These results can be used as information in delivering lecture material.