The evolution of information technology has revolutionized how humans engage with the world, particularly within the gaming sector. This paper explores the utilization of the DistilBERT Multilingual Cased model for analyzing sentiments expressed in Genshin Impact game reviews. The research methodology encompasses gathering data from Google PlayStore and Apple AppStore, manually labeling data, preprocessing it, and employing the DistilBERT Multilingual Cased model for analysis. The model's performance is assessed using metrics such as accuracy, precision, recall, and f1-score. Findings reveal that the model effectively categorizes sentiment in reviews, achieving an overall accuracy of 82%. Precision, recall, and f1-score metrics consistently surpass 0.77 across all sentiment categories. This study concludes that the DistilBERT Multilingual Cased model shows promise as a valuable tool for multilingual sentiment analysis within the realm of game reviews.
Copyrights © 2024