Analisis Spasial Kepuasan Layanan Puskesmas Kota Denpasar Menggunakan TF-IDF dan SVM

Abstract

Primary healthcare services delivered through community health centers play a strategic role in urban health systems, making the evaluation of public satisfaction with service quality a crucial aspect of continuous improvement. This study aims to analyze satisfaction with community health center services in Denpasar City by integrating sentiment analysis of Google Maps reviews using the Term Frequency–Inverse Document Frequency method and the Support Vector Machine algorithm, combined with spatial analysis at the sub-district level. The dataset consists of 2,681 user reviews collected from 11 community health centers across four sub-districts. The results indicate that the unigram TF-IDF model with a Linear Support Vector Classifier achieves the best performance, with an accuracy of 88.1% and a macro f1-score of 0.8534. Spatial analysis reveals variations in service satisfaction among sub-districts, with North Denpasar showing the highest satisfaction level, while South Denpasar falls into the moderate category and requires priority improvements. Furthermore, dominant word analysis indicates that service quality, staff attitude, and time efficiency are the main factors influencing both satisfaction and dissatisfaction perceptions. This study concludes that integrating machine learning-based sentiment analysis with spatial analysis provides a comprehensive overview of community health center service quality and can support evidence-based policymaking for more targeted improvements in primary healthcare services