SAVis : a Learning Analytics Dashboard with Interactive Visualization and Machine Learning

A dashboard that provides a central location to monitor and analyze data is an efficient way to track multiple data sources. In the educational community, for example, using dashboards can be a straightforward introduction into the concepts of visual learning analytics. In this paper, the design and implementation of Student Activity Visualization (SAVis), a new Learning Analytics Dashboard (LAD) using interactive visualization and Machine Learning (ML) is presented and discussed. The design of the dashboard was directed towards answering a set of 22 pedagogical questions that teachers might want to investigate in an educational dataset. We evaluate SAVis with an educational dataset containing more than two million samples, including the learning behaviors of 6,423 students who used a web-based learning platform for one year. We show how SAVis can deliver relevant information to teachers and support them to interact with and analyze the students’ data to gain a better overview of students’ activities in terms of, for example, their performance in number of correct/incorrect answers per each topic