Review of EDM for Analyzing the Performance of Students in Educational Setting

Data Mining plays an importnt role in Education setor in which it extracts the useful pattern from data collected from education field. So educational data mining (EDM) is related to the use of data mining, machine learning and statistical analysis technique on education related data. EDM is used to analyze the academic performance of students, predicting the students' performance in MOOC, to identify the students” educational behavior, to predict the dropout student in online courses, etc. In this study, the systematic review of EDM is considered related to the data mining algorithms - Classification technique, Clustering technique, Association Rule Mining algorithms and other technique such as graph mining, process mining, sentiment analysis, etc. For review purpose, 49 articles from Elsevier Journal or Procedia in EDM were analysed. It is found that maximum EDM research is done by comparing various classification techniques and choosing the best among them. It is noted that accuracy is the most widely used performance metric for classification while support and confidence are the performance metric used for association rule algorithm. This study also focuses on future scope in this domain to help the researchers.