“…A revision of the current and future state of EDM can be seen in the review performed by Baker and Yacef (). The raw data generated by different virtual learning environments (VLEs) have been used for a wide variety of purposes, such as to predict course dropouts in massive open online courses (MOOCs; Kloft, Stiehler, Zheng, & Pinkwart, ) and also in the context of high school education (Márquez‐Vera et al ), as decision support for college admissions (Janecek & Haddawy, ), to predict if students are going to surpass a course or not (Delgado Calvo‐Flores, Gibaja Galindo, Pegalajar Jímenez, & Pérez Piñero, ), to predict the major that a student is going to pick, before the student actually enrolls in college courses (Pedro & Ocumpaugh, ), to adapt learning environments to students' cognitive styles (Guo & Zhang, ), to provide information about the performance of groups in collaborative learning environments (Perera, Kay, Koprinska, Yacef, & Zaane, ), for group formation depending on students' learning styles (Dwivedi & Bharadwaj, ), to support the recommendation elicitation process in online learning environments (Santos & Boticario, ), or to predict the score of a test before actually doing it (Feng, Heffernan, & Koedinger,(); Pardos, Gowda, Ryan, & Heffernan, ). These results have a direct impact on creating tools that can help to improve these learning experiences.…”