Analysis of Public Sentiment Against PPKM Policy on social media Twitter Using Naive Bayes Classifier (NBC) Method Naufal Zhafran Albaqi, Suyono, Dania Siregar
Statistics Department, FMIPA Universitas Negeri Jakarta, Indonesia
Abstract
The implementation of Community Restrictions (PPKM) is one of the policies of the Indonesian government in preventing the spread of Covid-19. The implementation of the PPKM policy that has been going on for a long time has generated many responses in the community. Twitter is one of the social media used by the public to respond to the implementation of the PPKM policy. In this study, an analysis of public sentiment will be carried out on the PPKM policy on Twitter social media using the Naive Bayes Classifier (NBC) method. In addition, an overview of topics that are often discussed on positive, negative, and neutral sentiments in the implementation of PPKM policies will also be seen. The NBC method is a classification method based on the application of Bayes^ theorem, this method was chosen because it is faster and very good for text classification. The results showed that the NBC method was able to obtain accuracy on training data ranging from 68% to 71%. Meanwhile, the accuracy rate on the test data is 71%. These results indicate that the Naive Bayes Classifier algorithm has a fairly good performance. On negative sentiment, topics that are often discussed are the extension of PPKM, naming leveled PPKM, road closures, limiting meal times, and implementation of a work from home system. Meanwhile, on positive sentiment, topics that are often discussed are the implementation of better health protocols, and vaccinations, as well as decreasing PPKM levels and decreasing Covid-19 confirmation cases
