Data mining and an application on ASnadolu University distance education system

Nowadays, like in the other fields of study where data mining is applied, large databases are also available in education, from which meaningful relationships could be discovered and useful knowledge extracted. Within the scope of this PhD thesis, data mining is conducted by conjoining data from different sources about the students in Anadolu University Distance Education System. A model has been prepared to predict the student performance which could contribute planning activities of Distance Education System and clustering study has been made to determine the profiles of the graduates. A prediction model has been proposed in which C5.0 decision tree algorithm is used for predicting student achievement. It has been foreseen that the system could be used for the prediction of student achievement by integrating the decision rules of the proposed model. In result of the clustering, five clusters have been obtained by the use of ?K-means? algorithm. According to these clusters, it has been seen that they verify the relationship between the computer usage and student achievement.In the result of studies performed on the limited data in distance education students? database, it has been proven that data mining is an important decision support tool for internet-based distance education systems.