<![CDATA[This study aims to analyze user sentiment towards the Lokapala game through reviews collected from the Google Play Store. Lokapala is a local MOBA game developed by Anantarupa Studios that integrates Indonesian cultural elements. A quantitative approach is employed using the Support Vector Machine (SVM) algorithm to classify user reviews into positive and negative sentiments. Data were collected using the google-play-scraper library and preprocessed through several stages, including cleaning, case folding, word normalization using kamuskatabaku.xlsx, tokenizing, stopword removal, and stemming with the Sastrawi library. Reviews were labeled based on user ratings and split into training and testing datasets. Model testing results show an accuracy of 83%, with the highest precision of 0.85 for the positive class, recall of 0.93, and f1-score of 0.89. Additionally, WordCloud visualization revealed frequently occurring words such as "bagus" (good), "main" (play), "tolong" (please), and "banget" (very), reflecting both praise and technical complaints from users. These findings demonstrate that SVM is effective for sentiment analysis of user reviews and can provide valuable insights for developers in improving the quality of local games.]]>
Copyrights © 2026
