E-Learning personalization can be a solution to accommodate the variation of students' type since e-Learning allows the learning process of each student not to interfere with each other. Many variables may affect students' condition and behavior throughout the semester. Hence, students' type may change over time and different from one subject to another subject. Accordingly, the personalization should also be dynamic towards the changes that occur. In this research, we analyzed the students' type by processing the access log available on Learning Management Systems (LMS) from time to time. The results of the student analysis then become the reference for learning personalization using ontology. By utilizing ontology, personalization was presented by linking the students' type with activities that match the topic. The proposed personalized learning was applied to the prototype LMS later for testing and evaluation. The evaluation results indicated that personalized learning affects significantly to increase learning activities.