Analysis of SWCET Student’s Results Using Educational Data Mining Techniques

Abstract

AbstractEducational data mining (EDM) is a process of determining meaningful and useful knowledge from large amounts of data that are dug up from educational locations. The purpose of EDM is to discuss tools, techniques, and design research for gleaning meaning from large data sets on the fly. EDM is a research area that applies to machine learning, data mining, and pattern recognition techniques. Predicting student performance is an application of EDM to analyse student results. The paper aims to predict students’ results and compare different classification algorithms, considering educational data from Shadan Women’s College of Engineering and Technology from the past four years in a directive to predict and analyse students’ performance by identifying different data mining techniques such as J48 and logistic regression algorithms using the WEKA tool. This paper will contribute to constructing classification models for the ‘SWCET’ data set consisting of 627 different instances with eight different attributes for the first year, second year 376 instances, third year 360 instances, and final year 344 instances. It evaluates and compares implementation results to improve prediction accuracy. This study's findings, in particular, provide more insight for evaluating student performance with 100% accuracy using J48 and the logistic regression algorithm, which yields 99.84%. The student's performance will be advantageous to developing the quality of education and knowledge to drop the failure rate. All these things will help to improve the quality of the college.KeywordsData miningEDMClassificationJ48Logistic regressionWeka tool