Analisis Sentimen Terhadap Komentar Video IShowSpeed Tour Indonesia Pada YouTube Menggunakan Metode SVM

Abstract

In September 2024, international influencer IShowSpeed's visit to Indonesia attracted public attention and triggered various reactions on social media, especially YouTube. The content in the form of tours and live broadcasts conducted by IShowSpeed generated various comments from users, ranging from positive, negative, to neutral sentiments. This study aims to analyze these sentiments using the Support Vector Machine (SVM) method with a linear kernel. A total of 43,778 comments were used in this study. The classification results showed an accuracy of 91.1%. For negative sentiment, precision 86%, recall 78%, and f1-score 82% were obtained. Neutral sentiment achieved 90% precision, 94% recall, and 92% f1-score. Meanwhile, positive sentiment obtained a precision, recall, and f1-score of 94% each. These findings show that the majority of user comments are positive, indicating that IShowSpeed and its content are well received by Indonesian audiences.