<![CDATA[Impatient behavior among gamers is often reflected in sharp and emotionally charged digital reviews, particularly in the use of community applications such as Discord. This study explores expressions of impatience through sentiment and topic analysis. By adopting the CRISP-DM framework, a total of 10,000 Indonesian-language reviews collected from the Google Play Store were analyzed. The analytical process begins with sentiment labeling using IndoBERT, followed by polarity classification using the Support Vector Machine (SVM) algorithm, and topic exploration through the Latent Dirichlet Allocation (LDA) method. The results indicate that 57.4% of the reviews express positive sentiment, primarily related to voice communication quality and community interaction features. In contrast, 42.6% of the negative reviews commonly convey frustration regarding login issues and verification processes. The SVM model optimized using Bayesian Optimization achieved an accuracy of 90.46%. This study highlights that Discord serves not only as a communication platform but also as a reflection of users’ high expectations for system speed and stability. The main contribution of this research lies in the integration of SVM–LDA methods within the CRISP-DM framework to better understand the digital behavior of Indonesian gamers. The practical implications of these findings provide strategic insights for developers to improve authentication reliability and community features in alignment with user characteristics.]]>
Copyrights © 2026
