Modern educational systems, especially those mediated by human–computer interaction systems such as web-based courses, learning content management systems, and adaptive and intelligent educational systems, make large data streams available by recording student actions at microgenetic [25, p. 100] timescales* for days, weeks, and months at a time. Educational data mining (EDM) [31] aims to exploit these data streams to provide feedback and guidance to students and teachers and to provide data for educational researchers to better understand the nature of teaching and learning.