Computational Sentiment Analysis of User Interactions on Live Streaming Platforms Using Artificial Expert Judgment

Abstrak

This study examines user interaction patterns on JKT48 Showroom live streaming using sentiment analysis on 14,857 interactions processed with a Large Language Model. The results show that most interactions are neutral (93.63%) with positive and negative sentiments appearing in smaller proportions. This suggests that user participation is mainly routine and reflects a form of “silent loyalty”. The findings highlight the importance of maintaining a stable user experience to support long term engagement.

Identifier

• DOI: https://doi.org/10.24167/sisforma.v13i1.15480
• OpenAlex: https://openalex.org/W7166553955