“…Considering that the student modeling process is itself a diagnosis problem (Yudelson, Medvedeva, & Crowley, 2008), it is not surprising to see machine learning mechanisms build in student modeling layers of many studies (Aslan & İnceoglu, 2007;Tsiriga & Virvou, 2004). There are four mains problems to be addressed when machine learning is used with a student modeling system: the need for a large amount of data (Beck & Woolf, 2000), computational complexity, the crucial need for tagged data, and the concept drift problem (Webb, Pazzani, & Billsus, 2001). …”