This study intends to make a machine learning algorithm to quantify sentiment analysis to examine the effectiveness of using Virtual YouTuber (VTuber) Shinna Faultline, introduced by BMKG Gowa, as an alternative medium for delivering disaster information to adolescents. The AI that included a scraper, which was written in VScode, uses a natural language processing module with an incorporated dataset and data module known as Vader_Lexicon. To measure public understanding, a quantitative approach with a pre-test and post-test design was conducted at a junior high school. The research instrument was a dichotomous questionnaire that included favorable and unfavorable items which were processed with a paired sample t test. In conclusion there is a strong correlation between the comment section sentiment analysis and teenager understanding of disaster mitigation where if the video has a bigger percentage of positive comments, the understanding of the material generally increases amongst the students.