<![CDATA[One of the digital health service applications that are widely used in Indonesia is Halodoc. User reviews serve as a crucial source of information for developers to improve service quality. This study implements Aspect-Based Sentiment Analysis (ABSA) using Support Vector Machine (SVM) to classify user reviews based on aspects and sentiment. Data was collected through web scraping from the Google Play Store and categorized into four main aspects: Application Performance, Service Quality, Feature Completeness, and Pricing. Sentiment classification includes positive, neutral, and negative categories. The results show that positive sentiment is the most dominant (41.1%), followed by neutral (28.5%) and negative (30.4%). The SVM model achieves the highest F1-score in positive sentiment (0.85), while the Pricing aspect has the lowest accuracy (F1-score 0.46). This study demonstrates that ABSA can help Halodoc better understand user satisfaction, especially in aspects that frequently receive negative reviews, such as service pricing and delayed medicine delivery.]]>
Copyrights © 2025
