Comment Dataset on Presidential 3rd Debate.

<p><em><span>The presidential election of the Republic of Indonesia (RI) some time ago was the main topic of public concern, especially at the third debate of the Indonesian presidential candidates which was aired on YouTube on January 7, 2024. At the event, there were many comments from YouTube users in the comment column that were interesting to analyze. This study aims to compare the performance of two classification algorithms in machine learning, namely na&iuml;ve bayes and support vector machine (SVM). The stages passed in this study are data collection, data preprocessing, model classification and evaluation results. The data used is comment data in the comment column in the YouTube application which is crawled using Netlytic which produces 2.440 comments for analysis. Both algorithms are used to classify sentiment into two classes, namely positive and negative. The result of its performance is that the accuracy of na&iuml;ve bayes is better than SVM for the dataset used, namely the accuracy of na&iuml;ve bayes is 94.85% and SVM is 84.92%. </span></em></p>