Twitter is a social media that is currently popular, where the public is free to comment and write anything. It is not uncommon for the public to comment with harsh words and even hate speech. The 2019 presidential election drew many comments, some praised, criticized and insulted. To be able to dig up information and classify a text, sentiment analysis is needed. In this study, sentiment analysis is a process of classifying textual documents into two classes, namely negative and positive sentiment classes. Opinion data were obtained from the Twitter social network in the form of tweets. The data used was 3337 tweets consisting of 80% training data and 20% training data. Training data is data with known sentiment. This study aims to determine whether a tweet is a positive or negative tweet conveyed on Twitter in Indonesian. The classification of tweet data uses the naïve Bayes classifier algorithm. The classification results of the test data show that the Naïve Bayes Classifier algorithm provides an accuracy value of 71%. The accuracy value for each sentiment is 71% for positive sentiment and 70% for negative sentiment