Implementasi Algoritma Clustering dan Classification dalam Data Mining: Systematic Literature Review terhadap Tren dan Tantangan Terkini

This study conducts a systematic literature review on the implementation of clustering and classification algorithms in data mining to identify methodological trends and contemporary challenges during the 2021-2025 period. The research methodology employs the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach. Analysis was performed on eight relevant studies from IEEE Xplore, ScienceDirect, Springer, and ACM Digital Library databases. Narrative synthesis was used to comprehensively organize research findings. The results demonstrate the dominance of classification algorithms at 50%, with Random Forest achieving optimal accuracy of 98.35% through Particle Swarm Optimization. Clustering techniques demonstrate effectiveness in data segmentation, with K-means producing optimal configuration through Davies-Bouldin Index of 0.47. Application domains are diversified with the healthcare sector dominating 37.5% of implementations. Applications include diabetes prediction and COVID-19 epidemiological analysis. Hybrid approaches integrate various techniques for comprehensive knowledge extraction, particularly in social media user behavior analytics. Major challenges include computational complexity, methodological transparency deficiency in 66.67% of studies, and algorithm scalability limitations. Practical implications indicate a paradigm transformation in organizational decision-making from reliance on subjective intuition toward objective data-based formulation. Business intelligence technology penetration reaches 31.18% for dashboards and 10.75% for clustering techniques in small and medium enterprise ecosystems, marking substantial evolution in contemporary managerial practices.