Prediction of Indonesian presidential candidates in 2024 using sentiment analysis and text search on Twitter

Abstract

This study aims to show whether Twitter analysis can predict and forecast candidates in the Indonesian presidential election in 2024. This study was conducted long before the election began. To reduce the gap and utopian attitude in analyzing, two forms of analysis were used at once, namely sentiment analysis and text searches on Twitter data. This study uses a quantitative approach with descriptive content analysis. The data was obtained from Twitter social media, with Twitter Search focusing on official accounts and topics surrounding the 2024 presidential election. The search and data collection first adjusted to the trend of poll results spread in online news. The trend resulting from the poll is used to adjust the names of candidates to be searched for on Twitter search. The analysis tool used also utilizes the Nvivo 12 Plus analysis software. This study succeeded in mapping out three potential candidates in the 2024 election, namely Anies Baswedan, Ganjar Pranowo, and Prabowo Subianto. The mapping of potential candidates also has correspondence with the results of opinion polls in newspapers. From these findings, the information and data on Twitter help make predictions and an alternative to using the poll method. The drawback of this study lies in the limited use of time, so it is recommended that further research be carried out to collect and analyze similar data regularly until the election period. This may indicate that Twitter can predict earlier or better than polls.