Sentiment Analysis of Presidential Debate Videos on YouTube in the 2024 Indonesian Presidential Elections
Ariadani Dwi Fathurahman, Dita Pramesti, Hanif Fakhrurroja · 2024
Indonesia'$s$presidential debate is one of the series before the 2024 presidential election to convey each candidate's vision and mission. As the debate was broadcasted on popular video streaming platforms such as YouTube, there were various public responses to the candidates and the debate in the video comments. This study will analyze the public's sentiment in the comments on the presidential debate video towards each candidate. This study will use the YouTube API to collect video comment data and will use the Support Vector Machine (SVM) algorithm as a method of classifying comments as either positive or negative sentiment. The best training and testing data split in the testing is with a ratio of 80:20. With this ratio, it will produce an accuracy score of 0.90 and an AUC value of 0.95 for the Anies dataset. In the Prabowo dataset, it produces an accuracy score of 0.85 and an AUC value of 0.92. In the Ganjar dataset, it produces an accuracy score of 0.88 and an AUC value of 0.92. Furthermore, based on the results of sentiment prediction analysis, it can be seen that positive sentiment in Anies contains support for the vision and mission, in Prabowo contains full support in the general election, and in Ganjar contains satisfaction with the track record. Meanwhile, in negative sentiment, it can be seen that in Anies, it contains distrust of Anies' plan; in Prabowo, it contains doubts about voting; in Ganjar, it contains dissatisfaction with the results of the debate.