<![CDATA[Tax increase policies often generate pros and cons among the public, especially when perceived as having an impact on increasing unemployment. This study aims to analyze public sentiment regarding the issue of tax increases impacting unemployment by utilizing Machine Learning classification methods, namely Support Vector Machine (SVM) and Multinomial Naive Bayes (MNB). The data used comes from social media platform X in the form of public opinions collected online and then categorized into three sentiments: positive, negative, and neutral, with a total of 1,000 sentiment data points. The analysis process included text preprocessing, feature extraction with TF-IDF, and classification using both methods. In the Test and Score algorithm, the SVM algorithm produced an AUC of 0.660, CA of 0.694, F1 of 0.569, and Recall of 0.694, while the MNB algorithm produced an AUC of 0.586, CA of 0.198, F1 of 0.105, and Recall of 0.198. The study concluded that Support Vector Machines (SVMs) had a higher level of accuracy than Multinominal Naïve Bayes in classifying public sentiment. The majority of public opinion tended to be negative, indicating concern about the impact of tax increases on the workforce. These findings provide important insights for policymakers to consider public perception when establishing future fiscal policy.]]>
Copyrights © 2025
