<![CDATA[Regulatory enforcement against illicit cosmetics in the Papua Special Autonomy regions remains constrained by profound geographical complexity and systemic data fragmentation. This study develops a robust Decision Support System (DSS) that bridges the gap between digital surveillance and legal accountability at BBPOM in Jayapura. By integrating the Simple Multi-Attribute Rating Technique (SMART) with a hybrid machine learning framework, the system operationalizes parameters derived from the Indonesian Health Law (No. 17 of 2023) and the Papua Special Autonomy Law (No. 2 of 2021). A high-fidelity dataset of 1,324 multi-source inspection records was utilized to train a hybrid ensemble architecture. Empirical results demonstrate that the Artificial Neural Network (ANN) achieves near-optimal performance, yielding a 99.62% accuracy and 99.90% F1-score, with a marginal error rate of 0.38%. The inclusion of PostGIS-driven spatial intelligence further enables real-time vulnerability mapping within a scalable Service-Oriented Architecture (SOA). Beyond its technical efficacy, this research contributes a novel paradigm for localized law enforcement, successfully unifying regional regulatory mandates with advanced predictive analytics to safeguard public health in marginalized frontiers.]]>
Copyrights © 2026
