Sentiment Analysis and Topic Modelling of Public Opinion on TikTok Regarding Palm Oil Plantation Expansion in Papua

Palm oil plantation expansion in Papua has triggered diverse public responses on TikTok. This study analyzes public sentiment and identifies discussion topics regarding this policy using a deep learning approach. Data comprises 3,999 comments collected from the official Kumparan news account concerning President Prabowo Subianto's speech on Papuan land opening. We utilized IndoBERT for sentiment analysis and BERTopic for topic modeling. The results indicate that negative sentiment predominates (39.31%), followed by neutral (38.51%) and positive (22.18%) sentiments. This negativity stems primarily from ecological concerns, specifically flood risks and deforestation, alongside economic stability issues. Furthermore, topic modeling reveals four primary themes: Political Actors, Environmental Issues, Economic Issues, and Emotional-Spiritual Reactions. The study highlights a disconnect between government development goals and public fears concerning environmental risks and household welfare. As a practical implication, these findings provide a direct empirical basis for policymakers. The government must prioritize ecological mitigation and economic stability measures when designing sustainable development strategies in Papua to effectively address digital public concerns.