Exploring Audience Responses To The Pelangi Di Mars Movie Trailer On Youtube

Abstract

This study examined audience responses toward the Pelangi di Mars movie trailer on YouTube using sentiment analysis and topic modelling approaches. The data were collected from the YouTube comment section on March 10, 2026, resulting in 3,006 comments used as the research dataset. The comments were processed through several preprocessing stages, including text cleaning, tokenization, stopword removal, stemming, and normalization. Sentiment analysis was conducted using the IndoBERT pre-trained model to classify comments into positive, neutral, and negative categories, while Latent Dirichlet Allocation (LDA) was applied to identify dominant discussion topics within audience comments. The findings show that positive sentiment dominates audience responses, accounting for 73.7% of the total comments, followed by 13.9% neutral comments and 12.4% negative comments. Topic modelling reveals five major discussion themes related to animation quality, cinema enthusiasm, Hollywood comparison, story appreciation, and the development of the Indonesian animation industry. In addition, the IndoBERT model achieves an accuracy score of 94%, indicating strong performance in classifying Indonesian-language YouTube comments. The study concludes that YouTube comments function as electronic word of mouth (e-WOM) that influences public perception and supports digital film promotion and audience engagement in the entertainment industry.