Recommender System for Predicting the Academic Status of Students of Bangladesh Using Artificial Neural Network

Abstract

Ehsan Shirzad

Today, information technology in different areas produces a large amount of data every day. In recent years, many researchers have started to extract knowledge from data to make data-based decisions and improve the quality of processes in various organizations. Educational data mining helps educational institutions to operate effectively and efficiently by leveraging data from all stakeholders. It can help students at risk; create recommendation systems and alert students at different levels. Recently, neural network has received much attention in the educational sector compared to other methods. In this article, a study has been done using neural network on the dataset of Bangladesh University. The evaluation criteria for performance comparison are MAE (mean absolute error) and MSE (mean square error). The MAE and MSE of the neural network in the first run are 0.1357 and 0.026123, respectively. With the neural network, a recommender system has been presented to predict the students' academic status, which considers the student's status in four states: "excellent", "good", "needs more effort" and "needs guidance".

PDF