ANALISIS SENTIMEN PENGGUNA MEDIA SOSIAL TWITTER TERHADAP WABAH COVID-19 DENGAN METODE NAIVE BAYES CLASSIFIER DAN SUPPORT VECTOR MACHINE

Abstract

Twitter is often used to express opinions about a topic or issue that is trending. In the early 2020 period in Indonesia, Twitter was enlivened by the issue of the COVID-19 virus caused by SARS-CoV-2. Many Twitter users have expressed their views on the COVID-19 issue, which has attracted the attention of several parties to be used as a reference in making new decisions or policies. Therefore, it is necessary to do a sentiment analysis to determine the polarity of the sentiments that are in the contents of the tweets. This study uses the Naïve Bayes Classifier (NBC) and Support Vector Machine (SVM) methods. with a total dataset of 1652 tweets. From the results of classification using the NBC method, the classification accuracy value is 78.3%. While the accuracy value obtained by the SVM method is 81.6%. While the results of the accuracy test using the Cross Validation method with 10 K-Fold CV results in an average accuracy value of the NBC method of 69.8% and an average accuracy value of the SVM method of 74.4%. It can be concluded that the SVM method is proven to have a higher accuracy value than the NBC method.