Application of the Naive Bayes Algorithm in Twitter Sentiment Analysis of 2024 Vice Presidential Candidate Gibran Rakabuming Raka using Rapidminer

Abstract

In the current era of digital democracy, social media sentiment analysis has become a relevant method for understanding public views of political figures. As one of the leading social media platforms, Twitter provides a public space for sharing opinions and expressions regarding political issues. This research aims to classify and measure the accuracy of people's responses to the positive and negative sides. Sentiment analysis was carried out using the Naïve Bayes method using a dataset of 3223 tweets. The final results of this research show that implementing the Naïve Bayes Method in sentiment analysis regarding political dynasty polemics, especially regarding the 2024 Cawapres Gibran Rakabuming Raka, provides an accuracy value of 82.19%. Of the 1696 negative and 112 positive sentiments predicted, there were 462 harmful and 953 positive predicted data. These results indicate that most public responses tend to be detrimental to the Constitutional Court's (MK) decision, which grants political legitimacy to Gibran Rakabuming Raka as the 2024 vice-presidential candidate.