Data Mining untuk Evaluasi Kualitas Layanan Persalinan: Studi Komparatif RapidMiner dan SPSS

Abstract

Improving the quality of maternity services is a strategic priority in the health system. With
the development of technology, data mining has become an effective approach to evaluate the
quality of big data-based services. This study aims to compare the performance of two data
mining tools—RapidMiner and SPSS—in analyzing labor service data to assess the
effectiveness, efficiency, and ease of interpretation of the analysis results. A quantitative
approach was used as a method with a comparative design of 500 childbirth data from hospitals.
Data were analyzed using RapidMiner with Decision Tree and K-Means algorithms, as well as
SPSS with logistic regression and correlation tests. The indicators assessed include prediction
accuracy, processing time, and ease of use. The results of the analysis showed that RapidMiner
achieved a prediction accuracy of 85.4% and was able to cluster with a silhouette coefficient of
0.65. The processing time is about 12 minutes. SPSS shows an accuracy of 81.2% with a faster
processing time of 8 minutes. Significant factors found include the mother's age, complications,
and type of delivery. RapidMiner excels in predictive analysis and big data processing, while
SPSS is more efficient for conventional statistical analysis. A combination of the two is
recommended to obtain more comprehensive service evaluation results. The integration of data
mining in health information systems needs to be strengthened to support data-based policies in
improving the quality of maternity services.