YouTube is evidence of the development of leading digital technology in the media and entertainment sector. YouTuber is a term for those who create content or broadcast live on YouTube media. Not all YouTubers are able to express themselves directly in the content they create. For this reason, Virtual YouTubers were created, which are 2-dimensional or 3-dimensional virtual characters created using computer software to make it easier for YouTubers to interact with viewers. This research aims to analyze the live broadcasts of VTuber HoloH3ro, namely Kaela Kovalskia, Vestia Zeta, and Kobo Kanaeru, which contain schemes during the live streaming. With a large amount of data for each live stream time, a Naïve Bayes algorithm is needed through the use of range and frequency features, so accuracy results will be obtained which will be tested using the Confusion Matrix. The dataset used was taken from the kaggle.com site regarding the scheme live stream schedule conducted by HoloH3ro.