<![CDATA[YouTube is a widely used social media platform for expressing opinions on various issues, including local cultural events like Pasola Wanukaka. However, the diverse nature of comments makes manual sentiment identification inefficient, necessitating automated sentiment classification methods. This study aims to analyze public sentiment regarding Pasola Wanukaka and compare the performance of Lexicon-Based and Naïve Bayes methods in sentiment classification. The study utilized 601 YouTube comments collected via the YouTube Data API v3; following data cleaning, 464 comments remained for analysis. The research process involved preprocessing, sentiment labeling using the Lexicon-Based method, TF-IDF weighting, and classification using the Multinomial Naïve Bayes algorithm. The results indicate that positive sentiment was dominant, accounting for 211 comments (45.47%), followed by neutral sentiment with 175 comments (37.72%) and negative sentiment with 78 comments (16.81%). Evaluation results show that the Naïve Bayes method achieved an accuracy of 65.59%, outperforming the Lexicon-Based method, which achieved 60.99%. Thus, the Naïve Bayes method demonstrated superior performance in classifying the sentiment of YouTube comments regarding Pasola Wanukaka.]]>
Copyrights © 2026
