Purpose: This study aims to analyze the sentiment of comments on YouTube videos discussing the movie Dirty Vote, which are divided into positive, negative, and neutral categories.

Research methodology: The method used is quantitative descriptive, namely content/text analysis of the number of comments and sentiment analysis of comments on the YouTube video content of Dirty Vote uploaded to the Dirty Vote YouTube channel.

Results: Results: Based on sentiment analysis of 32,209 comments on the Dirty Vote movie uploaded to YouTube between February 11 and 13, 2024, the sentiment distribution indicates a strong dominance of negative sentiment, with 18,978 comments (59%). This suggests that the movie triggered substantial criticism or strong emotional reactions from viewers. Positive sentiment reached 9,676 comments (30%), reflecting appreciations for the movie’s content, message, or relevance to political discourse. Meanwhile, 3,555 comments (11%) were categorized as neutral, generally consisting of descriptive statements, clarifications, or non-emotive responses.

Conclusions: The analysis of 32,209 YouTube comments on Dirty Vote shows dominant negative sentiment (59%), followed by positive (30%) and neutral (11%). This suggests the film strongly impacts audiences, generating both criticism and political awareness, and highlighting how political documentaries can spark debate and polarization.

Limitations: References related to sentiment analysis on social media in the perspective of political communication studies are still quite limited and lack depth.

Contribution: Enriching the body of research on political communication on social media, particularly YouTube, by using sentiment analysis methods through the BERT model.