SENTIMENT ANALYSIS OF INDONESIAN PUBLIC RESPONSE TO THE 12% VAT INCREASE USING THE NAÏVE BAYES METHOD

I Made Ary Widiantara Putra (Informatika, Institut Bisnis Dan Teknologi Indonesia, Denpasar, Bali, Indonesia)
Ketut Jaya Atmaja (Informatika, Institut Bisnis Dan Teknologi Indonesia, Denpasar, Bali, Indonesia)
Putu Surya Lesmana (Informatika, Institut Bisnis Dan Teknologi Indonesia, Denpasar, Bali, Indonesia)
Desak Made Dwi Utami Putra (Informatika, Institut Bisnis Dan Teknologi Indonesia, Denpasar, Bali, Indonesia)
I Gede Andika (Rekayasa Sistem Komputer, Institut Bisnis Dan Teknologi Indonesia, Denpasar, Bali, Indonesia)

Jurnal Krisnadana
Vol 4 No 3 (2025): Jurnal Krisnadana May - July 2025

The policy of increasing the Value Added Tax (VAT) to 12% implemented by the government has sparked various reactions from the public, both supportive and opposing. This study aims to analyze public sentiment toward this policy using two classification methods: Naïve Bayes and Multinomial Naïve Bayes. The data was obtained through crawling from social media platform X with a total of 3,024 comments that had undergone preprocessing stages. The analysis was conducted using TF-IDF weighting, and model evaluation was performed using a Confusion Matrix. The evaluation results showed that the Multinomial Naïve Bayes algorithm performed better with an accuracy of 75%, compared to Naïve Bayes which only reached 71%. The majority of public sentiment towards the 12% VAT increase was negative, driven by concerns over economic impacts such as rising prices of goods. This study is expected to provide input for the government in formulating more targeted public communication strategies related to fiscal policies.