<![CDATA[This study aims to map the characteristic of data mining implementation based on 37 journals reviewed systematically. The method employed was a Systematic Literature Review (SLR) with PRISMA approach, consisting of identification, screening, eligibility, and inclusion stages. The data were analyzed descriptively to identify methods, algorithms, data types, and implementation result of data mining across various domains. The findings reveal that the most dominant task was clustering, with 27 articles, while K-Means was the most frequently used algorithm, appearing in 26 studies. The most widely studied implementation domain was economics and business with 12 articles, followed by health with 7 articles, social and government with 6 articles, also tourism and education with 5 articles each. In addition to K-Means, other algorithms identified in the reviewed studies included Apriori, FP-Growth, K-NN, Decision Tree/C4.5, Naïve Bayes, K-Medoids, and U-KMeans. In terms of evaluation, classification studies commonly used accuracy, association rule studies relied on support and confidence, while clustering studies tended to use measures such as the Davies-Bouldin Index and silhouette score. This study concludes that K-Means remains the most dominant algorithm in data mining implementation. The result of this review are expected to serve as a reference for understanding trends in data mining implementation and for supporting more focused future research.]]>
Copyrights © 2026
