Heru Teguh Santoso Teguh

Implementasi Data Minning Clustering Dalam Mengelompokan Kasus Perceraian Yang Terjadi Di Provinsi Jawa Timur Menggunakan Algoritma K-Means

Introduction

Implementasi data minning clustering dalam mengelompokan kasus perceraian yang terjadi di provinsi jawa timur menggunakan algoritma k-means. Analisis perceraian di Jawa Timur menggunakan data mining K-Means mengidentifikasi 7 klaster pola perceraian unik. Beri wawasan untuk strategi pencegahan perceraian yang tepat sasaran.

Abstract

Fenomena perceraian di Provinsi Jawa Timur menunjukkan tren peningkatan setiap tahunnya, didorong oleh berbagai faktor kompleks seperti kondisi ekonomi, pendidikan, dan sosial budaya. Penelitian ini menggunakan pendekatan data mining dengan algoritma K-Means untuk mengelompokkan kasus perceraian berdasarkan kesamaan karakteristik. Data perceraian dari Badan Pusat Statistik tahun 2020-2022 dianalisis menggunakan metode clustering, dengan evaluasi kualitas hasil clustering dilakukan menggunakan Davies-Bouldin Index (DBI).Hasil penelitian menunjukkan bahwa model clustering dengan 7 klaster memberikan performa terbaik dengan nilai DBI terendah, yaitu 0,460. Setiap klaster merepresentasikan pola distribusi perceraian yang unik, yang dapat digunakan untuk mengidentifikasi kelompok masyarakat yang rentan terhadap perceraian. Temuan ini diharapkan dapat memberikan wawasan yang mendalam bagi pengambil kebijakan dalam merancang strategi pencegahan perceraian yang lebih tepat sasaran. Penelitian lebih lanjut disarankan untuk memperkaya analisis dengan variabel tambahan seperti kondisi ekonomi dan tingkat pendidikan, serta menggunakan data yang lebih terkini.

Review

This paper addresses the significant and growing issue of divorce rates in East Java province, employing a data mining approach to uncover underlying patterns. The authors utilize the K-Means clustering algorithm to group divorce cases based on shared characteristics, aiming to provide a clearer understanding of the multifaceted factors contributing to this social phenomenon. The research is timely and relevant, given the increasing trend of divorce driven by complex socio-economic and educational factors, making the identification of vulnerable populations a crucial endeavor for social policy. The methodology is clearly outlined, involving the application of K-Means to divorce data from the Badan Pusat Statistik (BPS) for the years 2020-2022. A key strength of the study lies in its use of the Davies-Bouldin Index (DBI) to rigorously evaluate clustering quality, which found that a 7-cluster model provided the optimal performance with a low DBI of 0.460. This result indicates distinct and well-separated clusters, each representing unique patterns of divorce distribution. These findings are highly valuable, offering specific insights that can empower policymakers to design more targeted and effective prevention strategies, moving beyond generalized interventions. While the study presents valuable insights, the authors themselves identify several avenues for future enhancement. The current analysis, though effective, would greatly benefit from the incorporation of additional granular variables such as specific economic conditions and detailed educational levels, which are acknowledged as complex contributing factors but not explicitly included in the clustering model. Furthermore, leveraging more current data would ensure the findings remain highly reflective of contemporary trends. Addressing these aspects in future research would undoubtedly enrich the analytical depth and practical applicability of the model, providing an even more robust foundation for intervention.

Full Text

Comments

You need to be logged in to post a comment.