Abstract: This paper proposes to apply Data Mining Techniques to predict the studentsí failure on real time data of school or graduating students. Experiment attempts the detection of studentsí failure to improve their academic performance and to prevent them dropping out. Research has been done on assessing studentsí failure based on various attributes. In this experiment, 11 best attributes has been selected. Different approaches have been applied to resolve the problem of high dimensionality and using classification algorithm on Engineering studentsí previous and present education information to generate the model and this model can be used to detect studentsí academic failure. The results are compared and presented.

Keywords: Educational Data Mining (EDM), Academic failure, Classification, Prediction, Decision tree, Induction Rule.