Indonesian User Sentiment on Mobile Legends Bang Bang via Twitter

This study examines public sentiment toward Mobile Legends: Bang Bang (MLBB) in Indonesia using Natural Language Processing (NLP). A total of 10,363 Indonesian-language tweets posted between July 2020 and December 2024 were collected from X (formerly Twitter) and processed using Python. Sentiment classification was performed using a pre-trained Indonesian RoBERTa model to categorize tweets as positive, negative, or neutral. The results show a balanced sentiment distribution: 40.03% positive, 38.55% negative, and 21.42% neutral. Positive sentiment is driven by visuals, competitive gameplay, and community interaction, while negative sentiment is linked to technical issues, fairness problems, and monetization concerns. Keyword analysis highlights recurring topics that influence player engagement. The findings offer insights for MLBB developers and esports stakeholders to enhance user experience and address community concerns. This study contributes to gaming research by applying NLP to large-scale social media data in the Indonesian context.