Prediksi Tingkat Kepuasan dalam Pembelajaran Daring Menggunakan Algoritma Naïve Bayes

Abstract

The growth of learning at this time is influenced by advances in data and communication technology. One of the data technologies that functioned in the world of learning during the COVID-19 pandemic was online education. Online education is used as a liaison between lecturers and students in an internet network that can be accessed at any time. The online media used are Whatsapp, Google Classroom, Google Meet, Cloud x and the Zoom application. This research aims to predict the level of student satisfaction in online education as well as to distribute donations to large academies in making policies related to improving the quality of education online. The information used was obtained by distributing questionnaires to 110 students of the 2020/2021 class. The parameters in the questionnaire are lecturer communication, online education atmosphere, student evaluation, module delivery. Naïve Bayes is a prediction method for finding simple probabilities based on the Bayes theorem with a strong assumption of independence. Rapid Miner is one of the tools used for testing information and viewing the results of accuracy based on revolutionary information. The results of testing using 80 training information and 30 testing information display an accuracy of 100%. There were 3 respondents who reported dissatisfaction and 27 respondents reported being satisfied with online education. On the dissatisfied prediction, the precision class has a value of 100%, on the other hand, the prediction of being satisfied is 100%, and the class recall of true, not satisfied, has a value of 100%, whereas the class recall of true is satisfied to have 100%.