Social media is a medium used by users to introduce themselves, interact, collaborate, and share information with other users using the internet. One of the popular social media platforms in Indonesia is Twitter. Twitter is a social media that generally functions as a sender of messages which are usually referred to as tweets or tweets. One of the topics that has been widely discussed is the Policy on Enforcement of Restrictions on Community Activities (PPKM), due to the impact of an increase in cases due to the emergence of a new variant of COVID, namely the Omicron. One of the aims of this study is to find out the results of sentiment analysis regarding public opinion on the imposition of restrictions on community activities using the Naive Bayes method. There techniques machine learning for sentiment analysis, one of which is the Naive Bayes classifier, which is a machine learning technique based on probabilistic. NBC is a simple but very accurate and effective text classification method whose classification is heavily influenced by the training data process. The data used is taken via Twitter with 1594 tweets. The data set will be divided into training data and testing data by comparing 90% training and 10% testing. So, the details of the distribution of the data used in this study are 1594 tweets as training data and 160 tweets as test data. The NBC process crawling data pre-processing, data sharing, data labelling Bayes model naive classification, training data classification. The results of the analysis of public opinion sentiment regarding the imposition of restrictions on community activities using the Naive Bayes obtained a sentiment value of 71% sentiment negatif and 29% sentiment positif, accuracy value of 0.84, F1-Score 0.84, precision is 0.85, and recall is 0,84 .