Educational data mining: Methods and applications

Abstract

Educational data mining is a rapidly growing field that applies various statistical and data mining techniques to analyze educational data. This paper provides a general review of the literature on educational data mining, focusing on the methods and applications. Methods used in education data mining include classification and clustering. A classification algorithm is a supervised learning technique that seeks to categorize a given set of data objects into specified categories, build a classification model using the input data that already exists, and then apply the model to categorize new data items. The Naive Bayes, Decision Tree, Neural Network, and K-Nearest Neighbors have commonly employed classification algorithms in educational data mining. Clustering is unsupervised learning, whose objective is to divide a collection of data objects into various groups, where samples within a cluster exhibit a high level of resemblance and those between clusters are dissimilar. In educational data mining, the K-means Clustering Algorithm, Grid-Based Clustering, and Hierarchical Clustering are common clustering techniques. Those data mining algorithms are used in education such as student behavior prediction, student bad behavior detection, and student grouping. Overall, this research demonstrates that education data mining has a significant potential to improve educational programmers and student results. To solve the legal and privacy issues associated with the collecting and use of educational data, however, more research and solutions are required.