After the news of Kaesang's appointment as the Chairman of the Indonesian Solidarity Party (PSI), various speculations emerged on social media, particularly on Twitter (X). This study aims to classify sentiments regarding Kaesang's appointment as PSI Chairman using the Naïve Bayes algorithm optimized with Particle Swarm Optimization (PSO). The data used in this study consists tweets about Kaesang and tweets related to COVID-19. The text preprocessing process includes cleaning, case folding, tokenizing, stemming, and stopword removal. TF-IDF is used to represent words in vector form. In the initial experiment, Naïve Bayes performed classification using Kaesang data combined with COVID-19 data, with 300 data points for each label. Particle Swarm Optimization was used to improve the performance of the Naïve Bayes algorithm. The experiment results showed that the model tested with test data achieved the highest f1-score of 50%.