<![CDATA[This research aims to analyze the sentiment of YouTube comments on the music video "Horizon" by Jung Jaehyun by applying the Naïve Bayes and Support Vector Machine (SVM). As a global phenomenon, K-pop serves as an intriguing subject for understanding interaction patterns and fan opinions on social media platforms, particularly YouTube. A total of 2,391 Indonesian-language comments were collected using the YouTube API and processed through preprocessing stages such as data cleaning, tokenization, normalization, and the removal of common stopwords. After manually labeling the comments for positive and negative sentiments, the data was analyzed using the Naïve Bayes algorithm, known for its simplicity, speed, and effectiveness with small datasets, and compared with SVM equipped with a linear kernel. The study found that while SVM with a linear kernel achieved the highest accuracy of 98% and excelled in handling imbalanced data, Naïve Bayes still delivered competitive results with an accuracy of 97%. The advantages of Naïve Bayes, including ease of implementation, computational efficiency, and performance on small datasets, make it an effective choice for similar sentiment analysis cases. Both algorithms demonstrated good performance in predicting sentiments, as shown in their confusion matrices, although challenges persisted with the negative class. This research contributes to sentiment analysis methodologies by highlighting that Naïve Bayes is an efficient and relevant algorithm for preliminary exploration, while SVM is more reliable for performance optimization on complex datasets. The findings are particularly relevant to the music industry in understanding fan sentiment as an indicator of success.]]>
Copyrights © 2025
