<![CDATA[The hashtag #KaburAjaDulu, which went viral on platform X, reflects the concerns of Indonesian society—particularly among younger generations—regarding domestic social and economic pressures, as well as an increasing interest in migrating to Japan. This phenomenon illustrates the complexity of digital public opinion, yet few studies have specifically compared the effectiveness of different sentiment analysis algorithms within this context. Therefore, this study aims to analyze and compare public sentiment toward the #KaburAjaDulu hashtag, particularly about migration to Japan, using a Natural Language Processing (NLP) approach with three sentiment analysis algorithms: VADER, TextBlob, and BERT. A total of 1000 tweets were collected using scraping techniques, and after preprocessing, 967 tweets were included in the analysis. Sentiments were categorized into three classes: positive, negative, and neutral. The results show that VADER and TextBlob tend to classify tweets as neutral or positive, while BERT reveals a dominant negative sentiment of 52.3%. These findings suggest that BERT is more sensitive to context and implicit sentiment in the informal Indonesian language. This study highlights the importance of selecting appropriate algorithms for social media sentiment analysis and contributes to a deeper understanding of digital migration aspirations within Indonesian society.]]>
Copyrights © 2025
