Social media as a space for information and thought exchange has many users who have outsized influence towards other users. Users with such influence are known by the term 'influencers'. Their influence on social media mainly conveyed in the content they share, either in texts or images form. Such influence become an important aspect of societal life since majority of citizens are now social media user. This research analyzes best classification method to predict the sentiment contained in the response to contents shared by Indonesian influencer Fiersa Besari and Keanu in Twitter. The data used are tweets and comments in Bahasa Indonesia, gathered from Twitter API and cleaned up. The final dataset consists of 3,243 tweets with manual labeling of sentiment. The classification algorithm considered in this research are Naïve Bayes, Support Vector Machine (SVM), Logistic Regression, and Decision Tree. The main result of this research is that Naïve Bayes classification method has the best F-Score performance compared with other methods, with TP Rate for Fiersa Besari 88% and Keanu 86% & TN Rate for Fiersa Besari 76% and Keanu 60%.