A Student Success Prediction System for Open Education Zehra Kamisli Ozturk Gurkan Ozturk, Gurhan Ceylan, Sinan Aydin Studies on educational data mining (EDM) have increased in recent years, particularly due to the large number of students and their features. EDM pursues to find out patterns and make predictions that characterize learners’ behaviors and achievements, domain knowledge content, assessments, educational functionalities, and applications. The open and distance education systems have terabytes of data related to the students and graduates. These data have been serving to make strategic and operational decisions such as on location, on the numbers and capacities of the offices, and on the number of the books to be printed. In this study, we consider Anadolu University, third mega university of the world and has approximately two millions of students and more than two millions of graduates. For this study, Business Administration Program of Open Education Faculty is selected as a case. In order to calculate student success of this program, we use patterns related to students’ previous grades. The classification problems formed by these patterns are solved with mainstream classifiers. Experimental results show that student success can be predicted with at least %70 accuracy. The obtained classifiers will serve as a recommendation system for students who want to select courses before the semester registration. Keywords: education Data mining, classification, open