Sentiment Analysis of 2024 Presidential Candidates in Indonesia: Statistical Descriptive and Logistic Regression Approach

Sentiment analysis is a technique for analyzing public sentiment or a particular segment of sentiment toward an object. One of the implementations of sentiment analysis is toward potential presidential and vice-presidential candidates in the upcoming 2024 Indonesian Presidential Election (Pilpres). This upcoming Presidential Election is a topic that is frequently discussed in Indonesian politics, by voters, various political parties, and people’s community daily conversation anywhere. Many prospective candidates are rumored to be supported by various political parties to become their presidential and vice-presidential candidates, without any concern about their reputation and sentiment among targeted voters. However, the sentiment towards these candidates was not always clear, whether it leans towards positive or negative. To address this issue, this sentiment analysis uses Logistic Regression and Descriptive Statistics to do classification and analysis of the sentiment of tweets regarding each of the prospected candidates. The result of the sentiment analysis was a polarity score of tweets for these prospective candidates. The Logistic Regression method has an overall average accuracy score of 79% on all five prospected 2024 presidential candidate datasets. The highest accuracy is obtained at 93% on the Prabowo Subianto dataset and the lowest at 70% on the Anies Baswedan and Muhaimin Iskandar datasets. This analysis also provides valuable insights into the public sentiment toward potential candidates for the 2024 Presidential Election in Indonesia.