Implementation of the Support Vector Machine (SVM) Algorithm on Sentiment Analysis of Public Opinion on The Prohibition of the use of Syrupy Drugs for Kidney Health

In 2022, the Indonesian Ministry of Health reported several cases of pediatric acute renal failure (GGAPA), which resulted in a mortality rate of 59%, mainly among children aged between 1-5 years. The main causes were identified by Health Minister Budi Gunadi Sadikin as the three solvents ethylene glycol (EG), diethylene glycol (DEG), and ethylene glycol butyl ether (EGBE). In response, the government implemented restrictions on the consumption of these condensed substances, which led to mixed public reactions observed in the YouTube comments section. The purpose of this study is to evaluate public opinions on the syrup ban for kidney health. The comments will be classified using the Support Vector Machine (SVM) method, and the most effective kernel among linear, sigmoid, polynomial, and RBF will be identified. Data was collected through web scraping with 5000 initial data, and after preprocessing, 4794 data were processed. The analysis results show that the linear kernel has the highest accuracy of 75.63%, followed by the sigmoid kernel 75.29%, RBF 74.79%, and polynomial 71.09%. While the K-Fold Cross Validation test with a value of k = 10, produced a value of 74.64% for the linear kernel. This research concludes that the Support Vector Machine (SVM) algorithm with a linear kernel achieves the highest accuracy in sentiment analysis.