Performance Evaluation of College Students’ Google Classroom Engagement Using Data Mining Techniques

Abstract

The purpose of this study is to assess how well college students use Google Classroom as a useful and informative teaching and learning tool. The survey method was utilized in the study to measure student involvement in Google Classroom. This study's sample population included 292 college students from Northern Negros State College of Science and Technology. Algorithms such as Random Forest (RF), C4.5, and Naive Bayes (NB) were utilized with three of the most crucial techniques, such as 60% split, training set, and 8-fold cross-validation, for performing analysis on the student data. After analyzing different metrics for performance (Correctly Classified Instances, FP Rate, ROC Area, F-Measure, TP Rate, Recall, Precision, Time taken to build model, Mean Absolute Error, Root Mean Squared Error, Root Relative Squared Error, Relative Absolute Error) by various algorithms for data mining, the researchers determined which algorithm performs better than others on the student dataset gathered, allowing the researchers to make a recommendation for future improvement in students' Google Classroom engagement.