<![CDATA[Violence against women and children in Nagekeo Regency is a crucial social issue requiring targeted intervention. The Department of PMD-P3A faces challenges in analyzing regional vulnerability, which has historically been manual and subjective. This research aims to develop a web-based vulnerability grouping system implementing the K-Means++ Clustering method. This method was strategically selected for its ability to optimize initial centroid selection through distance probability calculations, resulting in more stable and accurate clustering compared to the standard K-Means algorithm. The system was developed using the Laravel framework and MySQL database, utilizing historical data from 2020 to 2025. The clustering process is based on two key parameters: Type of Violence and Place of Occurrence, mapping regions into three levels: Highly Vulnerable, Vulnerable, and Non-Vulnerable. The results demonstrate excellent system performance with a Silhouette Score of 0.6633 and a Davies-Bouldin Index (DBI) of 0.4520, indicating a solid and optimally separated cluster structure. Beyond statistical data, the system provides interactive digital mapping visualizations. This implementation is expected to serve as a decision-support tool for the local government in formulating more effective and efficient social protection policies in Nagekeo Regency.]]>
Copyrights © 2026
