In recent years, microblogging platforms have emerged as highly popular communication tools among internet users. Microblogging refers to a form of social media that enables users to share short messages—typically limited to fewer than 200 characters—containing opinions, comments, or news. One widely used microblogging service is Twitter, which attracts users from diverse backgrounds. Following the Indonesian presidential election in February 2024, numerous trending tweets expressed various opinions about Prabowo Subianto, the 8th President of Indonesia. These tweets reflected a range of sentiments, including positive, negative, and neutral viewpoints. Based on this, a sentiment analysis system is needed to efficiently analyze and classify public responses on Twitter according to sentiment categories: positive, negative, or neutral. To perform this classification, an appropriate algorithm is required. One suitable method is the Naïve Bayes classifier, a probabilistic algorithm that categorizes data by computing probability values. This method is well-suited for text classification tasks such as sentiment analysis of tweets. In this study, a dataset comprising tweets was used, consisting of 96 entries for training and 150 entries for testing, totaling 246 data points. The Naïve Bayes classifier achieved an accuracy rate of 84% using this dataset.
Sulaiman et al. (Mon,) studied this question.