Modeling and Visualizing Student Flow

In this work, we present a data science system to model and visualize student flow patterns based on electronic student data of a university. Our system is called eCamp. The datasets used by eCamp were previously disconnected and only maintained and accessed in a siloed manner by independent campus offices. At a campus-level, our models and visualization show how students make choices among hundreds of potential majors, as students gradually progress towards their sophomore, junior, and senior year. At a department-level, the student flow patterns revealed by eCamp show how each course plays a different role within a curriculum. eCamp further dives down to the granularity of the exact classes offered in each semester. At that level, eCamp shows how students navigate from one set of classes in one semester to another set in a subsequent semester. Previously, comprehensive information about student progression patterns at all of these level was simply unavailable. To that end, we also demonstrate how insights into such student flow patterns can support analytical tasks involving student outcomes, student retention, and curriculum design.