ABSTRACT
Education domain generates huge amounts of data, especially in online teaching and learning. Data analysis enables detecting patterns of students' learning behavior which leads to personalized attention and adaptive feedback. Learning management system (LMS) data are in the focus of this paper. The LMS logs store information about students' login frequency, time of visits, number of downloading different resources, time and frequency of various activities. Within this research, log file analysis is performedfrom a Business decision making course at the University of Zagreb. Raw LMS data were extracted from Moodle, data was prepared, explained and explored in order to detect patterns in student's behavior in the online environment. Research results provide a basis for teachers' interventions on one hand, and serve as an input into predictive models' development, on the other hand.