Sentiment Analysis of Personal Data Protection and Privacy Law in Indonesia Using Multinomial Naive Bayes Algorithm

The protection of personal data and individual privacy is critical in today's digital world. Recently, the number of digital privacy violations and personal data breaches in Indonesia has increased significantly. Therefore, through the Personal Data Protection Law (PDP Law) No. 27 of 2022, the Indonesian government encourages all parties, individuals, and industries to actively protect privacy and personal data. However, the understanding and perception of the public regarding this matter are very diverse. Some layers of society believe that personal data and privacy protection are less important, while others think that enforcing personal data protection is crucial to minimize cybercrime. This research aims to analyze public sentiment regarding PDP Law No. 27 of 2022 based on comments generated from the $\mathbf{X}$ social media. Multinomial Naive Bayes is adopted as the method to analyze public sentiment. The method is prominent in analyzing textual classification and is widely adopted for the classification of categorical text data. The evaluation results of automatic labeling show that the model achieves 74 % accuracy with precision, recall, and f1-score values of $86 \%, 78 \%$, and 82 % for negative sentiment, and $44 \%, 59 \%$, and 50 % for positive sentiment, respectively. Meanwhile, the evaluation results from manual labeling show that the model achieves 72 % accuracy with precision, recall, and f1-score values of $50 \%, 2 \%$, and 3 % for negative sentiment, respectively, and $\mathbf{7 2 \%}$, $\mathbf{9 9 \%}$, and $\mathbf{8 4 \%}$ for positive sentiment. The model is also implemented in a web application that can accept user input and check its sentiment.