<![CDATA[The transformation of application-based public services requires the government to provide reliable, usable, and trusted digital channels. However, evaluations of government applications often stop at rating scores or general sentiment classification, so they do not identify the service aspects that generate satisfaction or complaints. This study aims to evaluate Indonesian government applications through aspect-based sentiment analysis of Google Play user reviews. The data source is the IGAR: Indonesian Government App Review Dataset, which covers six applications, namely JMO, SATUSEHAT, Mobile JKN, MyPertamina, KAI, and BMKG. After removing 26 reviews without content, 617,696 reviews were analyzed. Sentiment labels followed rating-based labels, while service aspects were extracted using dictionary-based matching validated on 200 review samples. The results show that positive sentiment dominates at 54.47%, negative sentiment accounts for 39.97%, and neutral sentiment accounts for 5.56%. At the aspect level, system performance and account access show the highest negative proportions, at 82.49% and 81.69%, respectively. These findings indicate that government application improvement should prioritize system stability, authentication, data quality, and clearer service flows as the foundation of public digital information services.]]>
Copyrights © 2026
