Predicting Academic Performance through Data Mining: A Systematic Literature

Abstract

The main objective of this work is to make a systematic review of the literature on the prediction of the academic performance of university students by applying data mining techniques. For this purpose, an exhaustive search was carried out and after the analysis of the documentation collected, aspects such as: methodology, attributes, selection algorithms, techniques, tools, and metrics were considered, which served as the basis for the elaboration of this document. The results of the study showed that the most used methodology is KDD(database knowledge extraction), the most important attribute to achieve prediction is CGPA(academic performance), the most commonly used variable selection algorithm is InfoGain-AttributeEval, among the most efficient techniques are Naïve Bayes, Neural Networks (MLP) and Decision Tree (J48), the most used tools for the development of the models is the Weka software and finally the metrics necessary to determine the effectiveness of the model were Precision and Recall.