www.ijecs.in International Journal Of Engineering And Computer Science ISSN:2319-7242 Volume 3 Issue 12 December 2014, Page No. 9395-9398 Samira Talebi 1 IJECS Volume 3 Issue 12 December, 2014 Page No.9395-9398 Page 9395 Using Educational Data Mining (EDM) to Prediction and Classify Students Samira Talebi 1 ,Ali Asghar Sayficar 2 1 Islamic Azad University Garmsar Branch, Department ofInformation Technology, University Square, Student Street, Iran samiratlb86@gmail.com 2 Islamic Azad University Garmsar Branch, Department ofInformation Technology, University Square, Student Street, Iran a_sayficar@yahoo.com Abstract:The aim of this paper is to predict the students’ academic performance. It is useful for identifying weak students at an earl ier stage. In this study, we used WEKA open source data mining tool to analyze attributes for predicting students’ academic perfo rmance. The data set comprised of 180 student records and 21attributes of students registered between year 2010 and 2013. We chosethem from FERDOWSIUniversity of Mashhad .We applied the data set to four classifiers (Naive Bayes, LBR,NBTree, Best -First Decision Tree) and obtained the accuracy of predicting the students’ performance into either successful or unsuccessful class. The student's academic performance can be predicted by using past experience knowledge discovered from the existing database. A cross-validation with 10 folds was used to evaluate the prediction accuracy. The result showed that Naive Bayes classifier scored the higher percentage of prediction F- Measure of 83.9%. Keywords:Data Mining, Prediction, Average, Attributes for predicting students, Educational Data Mining (EDM) 1.Introduction Classification and prediction are of highimportance in data mining techniques and usedin many fields. Recently, researchers haveutilized machine learning in order to makewise career decisions. It is useful for both thestudents and the instructors getting better intheir performances. We got our dataset fromthe Information system of the biggest virtualuniversity of Iran. We decided to extract theattributes that have significant contribution tothe prediction of academic performance. Theprediction can be done by using data miningtools such asWeka software. 2. Methodology Many studies were undertaken in order to explain the academic performance or to predict the success or the failure (Kotsiantis et al., 2003; Chamillard,2006;Minaei- Bidgoli et al., 2003;Merceron and Yacef, 2005; Romero etal., 2008;Superby et al.,2006;Vandamme et al., 2007;Ardila, 2001; Gallagher, 1996; King,2000;Minnaert and Janssen, 1999;Parmentier, 1994.)they highlighted a series of explanatory factors associated to the student. We first considered a set of attributes to betaken into account based on a model used byParmentier (1994). Secondly, we created aquestionnaire allowing us to collect a largeamount of interesting information on a certainnumber of students. We distributed thisquestionnaire by paper to students in theFERDOWSI University of Mashhad. We used WEKA open source data mining. Itsupports many machine learning algorithmsand data processing tools. In the datapreprocessing step, we collected 205 records of students admitted from year 2010 to 2013 atthe FERDOWSIUniversity of Mashhad.According to the total average, thestudents were classified into four classes: Grade A (Total Average>=17), Grade B (15=<Total Average< 17), Grade C (13=<Total Average< 15),