PT PAL Indonesia is one of the state-owned enterprises engaged in the shipbuilding industry which has business advantages in shipbuilding and shipbuilding capabilities. Being a fairly large company, PT PAL gets opinions from the public regarding the performance and services provided. Therefore, a sentiment analysis was carried out on public opinion on Twitter social media using data that had been collected into a dataset and processed using Rapidminer tools. This study uses the Naïve Bayes, K-NN and Decision Tree methods to make comparisons by looking at the level of accuracy of the three methods used. The results of the study show that the Naïve Bayes method has an accuracy rate of 84.08% with class precision for pred. positive is 83.65%, pred. Neutral is 97.06%, pred. negative 100%, K-NN method is 83.38% with class precision for pred. positive is 83.05%, pred. Neutral is 96.43%, pred. negative 0.0% and the Decision Tree method is 81.09% with class precision for pred. positive is 81.09%, pred. Neutral is 0.0%, pred. negative 0.0%. The results of this study can show that the Naïve Bayes method has a higher accuracy rate than other methods used with an accuracy rate of 84.08%