This paper investigates means to visualize and classify patterns of study of a college math learning environment. We gathered logs of learner interactions with a drill and practice learning environment in college mathematics. Detailed logs of student usage was gathered for four months. Student activity sessions are extracted from the logs and clustered in three categories. Visualization of clusters allows a clear and intuitive interpretation of the activities within the clustered sessions. The

more »

... hree clusters are further used to visualize the global activity of the 69 participating students, which would otherwise be difficult to grasp without such means to extract patterns of use. The results reveal highly distinct patterns. In particular, they reveal an unexpected and substantial amount of navigation through exercises and notes without students actually trying the exercises themselves. This combination of clustering and visualization can prove useful to learning environments designers who need to better understand how their application software are used in practice by learners.