eMineProve: Educational Data Mining for Predicting Performance Improvement Using Classification Method

• Published in: IOP conference series. Materials Science and Engineering Vol. 649; no. 1; pp. 12018 - 12025
• Main Authors: Rosado, Jovel T., Payne, Angelica P., Rebong, Corazon B.
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.10.2019
• Subjects: Academic achievement; Classification; Colleges & universities; Data analysis; Data conversion; Data mining; Decision analysis; Market strategy; Marketing; Middle schools; Model accuracy; Performance prediction; Prediction models; Secondary school students; Secondary schools; Students
• ISSN: 1757-8981; 1757-899X
• DOI: 10.1088/1757-899X/649/1/012018

Today, Data mining has become a universal tool for converting data into useful information and knowledge. Student's academic performance is a crucial factor in building their future [3]. In this study, the classification method is selected to be applied on the students' data. Classification is a form of data analysis that extracts models describing data classes [4]. The data is collected from the Basic Education Department of the University of Perpetual Help System Laguna. The raw data set is a collection of 4, 250 data accumulated over two academic years regarding the basic information of the Grade 7 Junior High School students. The data mining technique applied was classification using Naïve Bayes. The data mining tool used to process the data into useful knowledge is RapidMiner. The main objective of this study is to predict the performance improvement of Grade 7 Junior High School students for A.Y. 2016-2017 and 2017-2018 using classification. It specifically aims to identify the general average of the Grade 7 Junior High School students when grouped according to gender, identify who perform better between male and female, identify the subject on which the students excel most, identify the subject on which students have difficulty, identify who performs best when group according to last school attended, identify the academic performance of the students based on their parent's occupation, provide a predictive analysis of data to help the decision makers create a marketing strategy for those schools where only few students enrolled. Naïve Bayes produces accuracy 92.37% that shows it is possible to obtain a good prediction model based on the academic performance of the students. Based on the data for marketing strategy, the model has an accuracy of 30.97%.