Comparative Analysis of Community Sentiment Against the Implementation of Booster Vaccination in Indonesia Using the K-Nearest Neighbor and Naïve Bayes Classifier Methods

Abstract

Sentiment analysis is a person's opinion or view of a particular object that produces positive, negative, or neutral sentiments. The government's effort during the COVID-19 pandemic is to call for the implementation of a booster vaccination program to the public. Based on this, it produces several public sentiments, some of which are uploaded on the Twitter social media platform, which generate positive and negative sentiments. To find out the classification of public sentiment, the researchers carried out calculations using the K-Nearest Neighbor and Naïve Bayes Classifier methods. Based on the calculation results, it was found that the public sentiment was positive at 98% and negative at 2%. This means that the community is enthusiastic and supports the booster vaccination program. Then the comparison based on the calculation results, namely the K-Nearest Neighbor method with a K value of 3 resulting in an accuracy calculation of 97.33% and using the Naïve Bayes Classifier method, an accuracy calculation of 97.35% can be generated. So it can be seen that using the Naïve Bayes Classifier method has a higher accuracy than the K-Nearest Neighbor method.