Prediction of Student Satisfaction with Academic Services Using Naive Bayes Classifier

Academic service is an academic activity offered to a party (student) either directly or indirectly in the context of achieving academic goals. Student satisfaction as users of educational services is very important for the progress of a university. This is because student satisfaction will have an impact on their loyalty to the university. The prediction of student satisfaction is an important factor that determines the quality of a university. Data mining can optimize the process of finding information in large databases and finding previously unknown patterns. The method used is a prediction with a Naive Bayes Classifier using 85 datasets obtained from surveys using the SERVQUAL scale. There are 4 different experiments in datasets distribution for training and testing to find the best distribution. The best distribution result is 75% for training and 25% for testing with 95% accuracy. Research on predicting student satisfaction with academic services at the study program of PTIK of UNNES using the Naïve Bayes classification method produces a good model and website application that predicts student academic service satisfaction with an accuracy value of 95% and it is concluded to be a very good classification model.