With the rise of online sales transactions through e-commerce and social media, the impact of changing consumer behavior. This has resulted in many sentiment assessments from users of these expedition services. Expedition services are now one of the most popular services. One of the shipping service companies operating in Indonesia is J&T Express. Through social media, especially Twitter with the number of followers up to 154,439 and the number of tweets up to 103,100, a user can form an opinion on the performance of J&T Express and get data as many as 1694 tweets. Machine learning algorithms are needed to enable sentiment analysis to be classified. One of them is the Naive Bayes algorithm. Before running the classification process, a preprocessing process is needed so that the dataset can be recognized by the system. Based on the tests conducted, the results obtained accurasy of 84%, precision of 76%, and recall of 87%. The results of this research show that the data can be used as a basis for evaluating business decisions.