<![CDATA[Indonesian Language Subjects are generally regarded as easy lessons and do not need to be studied by students and society. Based on this, various learning problems arose involving instructors, Indonesian language subjects, students who received lessons, teaching methods, facilities, ways to obtain, and the objectives of Indonesian language learning (Moeljono, 1989). The difference between each student in different learning differences. This causes the teacher to have limitations in measuring the level of understanding of students. Then a system is needed to predict the level of understanding of students. This prediction uses the classification method with the Naive Bayes algorithm. The class that will be used in this study is that students understand, are quite understanding and lack understanding. In this study, the authors used the Information Gain (IG) feature selection. The selected feature will be processed with the Naive Bayes classification algorithm, then the accuracy will be seen if it is not maximized, then the previous feature selection process will be done again to get the desired verification. From the tests that have been conducted, the results obtained which have a Gain value of more than 0.2 have the largest rating, reaching 90%. The features chosen from 17 included features of family members, residence status, mother's work, caregivers, family support, joining extracurricular activities, repeating lessons at home, length of study at home, reading at home, reading time at home.]]>
Copyrights © 2019
