Analisis Data Sentimen Pemain Game Role-Playing Game (RPG) Honkai Star Rail dengan Algoritma Naive Bayes

In the rapidly evolving digital age, the gaming industry has experienced significant growth. One of the games that is currently popular is Honkai Star Rail. This research uses Honkai Star Rail game player reviews as the object of sentiment data research. However, the problem found in sentiment data analysis is the existence of reviews that provide positive or negative ratings as a response from players. There are situations where these labels do not fully reflect the true essence of whether the response is positive or negative, so it is necessary to analyze sentiment data. The purpose of this sentiment data research is to assess the performance of the naive bayes model in classifying sentiment by finding the best accuracy value and AUC value from 3 scenarios of sharing test data and data. The method in this study uses the Naive Bayes algorithm, this algorithm was chosen because it is suitable for classification problems. The test results with 3 dataset sharing scenarios (60:40, 70:30, and 80:20) show that the accuracy value reaches the highest value in scenario 2 (70:30) which is 86%. The precision value also reaches the highest value in scenario 2, which is 84%, the recall value reaches the highest value in scenario 2, which reaches 89%. And the highest AUC (Area Under the Curve) value is obtained from scenario 2 of 0.92 with the excellent classification category.