Viewer Perception Analysis of Video Collaboration of the Nihongo Mantappu YouTube Channel Using Aspect-Based Sentiment Analysis

YouTube is an online video-sharing platform, one of the YouTube channels with collaborative content with the highest number of subscribers in Indonesia is the Nihongo Mantappu channel. This study uses the aspect-based sentiment analysis method for content development to analyze audience perceptions of collaborative video comments from the Nihongo Mantappu channel. The aspect used in this study is the type of content consisting of Cooking, Challenge, Education, Mukbang, QnA, and Vlog content. The Decision Tree model, with a sentiment analysis accuracy rate of 96.08% and aspect-based settlement analysis of 98.36%, shows that the overall proportion of sentiment is dominated by neutral sentiment with a percentage of 60.5%, positive sentiment with a percentage of 23.8%, and negative sentiment with a percentage of 15.7
^%
• The results of this study can be used as a reference in developing collaborative video content on the Nihongo Mantappu channel.