<![CDATA[With the growing use of digital healthcare platforms such as Halodoc, maintaining consistent service quality that meets user expectations is essential. User reviews on platforms like Google Play provide valuable insights into user perceptions. This study aims to classify user sentiments toward Halodoc’s pharmacy services based on reviews obtained through web scraping from the Google Play Store. The analysis employs the pre-trained IndoBERT model to extract textual features, followed by sentiment classification using the Random Forest algorithm. This combination was selected for its efficiency with limited hardware resources and small dataset size. To enhance data diversity and minimize overfitting, simple augmentation methods such as random word deletion and synonym substitution were implemented. The expected outcomes include an effective sentiment classification model and visualizations of sentiment distributions (positive, negative, neutral). Furthermore, the study contributes to the development of sentiment analysis techniques for Indonesian-language data through an efficient and contextually relevant approach. The research outputs target publication in a nationally accredited (Sinta 4) journal and Intellectual Property Rights (IPR) registration. Ultimately, this study is expected to support the improvement of technology-based pharmacy services through the strategic application of machine learning.]]>
Copyrights © 2025
