The document discusses using data and analytics in online education. It notes that online learning is increasingly popular as people can learn on their own schedule and from anywhere. However, online education faces challenges in keeping each unique learner engaged. The document proposes addressing this by collecting detailed interaction data and using algorithms to provide personalized recommendations and guidance to students and teachers. It outlines an architecture that would log student and faculty data, process the logs in Hadoop, and power data-driven applications to improve instruction and learning outcomes. Examples discussed include faculty dashboards providing insights and an adaptive math tutor enhancing activities based on student performance data.