Abstract:
Student’s performance is an essential part in learning institutions. Predicting student’s performance becomes more challenging due to the large volume of data in educational databases. The adoption of the educational data mining by higher education as an analytical and decision making tool is offering new opportunities to predict student performance. The university management would like to know which features in the currently available data are the strongest predictors of university performance. In order to help the academic advisor to monitor the students’ performance in a systematic way by identifies those students which needed special attention to reduce failing ration and taking appropriate action for the next semester at a right time. To meet these objectives the researcher used CRISP-DM Methodology which governs by a series of stages. Starting by business understanding followed by data understanding, data preparation, modeling evaluation and deployment. Many experiments conducted to find out a model that could be useful for predicting students’ performance based on their social factors using decision tree (j48, random forest) and Bayesian classifiers (naïve Bayes, Bayes net) as classification techniques. The experimental results showed that J48 is the best algorithm for classification of data. It also showed that social factors have got significant influence over students’ performance
