<![CDATA[The geographic conditions of the archipelagic region within the working area of Loka POM Tanimbar, combined with limited access to transportation and infrastructure, have led to uneven coverage in food distribution facility inspections. Data from these inspections reveal that the risk to food safety remains relatively high, as supervision has not been fully risk-based or regionally targeted. As such, optimizing inspections is crucial for improving business compliance with regulations and enhancing public protection. This study aims to identify the vulnerability profile of non-compliant food distribution facilities at the sub-district level, analyze the characteristics of the inspection network, and provide actionable recommendations for optimizing food distribution facility inspections through spatial and social network analysis. Using a descriptive quantitative method and Geographic Information System (GIS) approach with QGIS, along with the SIPT database from BPOM, spatial analysis through Hot Spot Analysis (Getis-Ord Gi*) and Global Moran's I in ArcGIS identified patterns of vulnerability in food distribution. Social Network Analysis with Gephi assessed the inspection network structure. The results showed that inspections were focused in administrative and economic centers like Tanimbar Selatan and Moa Lakor, with 13 sub-districts remaining unreached, creating blind spots for non-compliant food circulation. Expired products were the predominant issue, indicating poor stock management and internal control among business operators. Spatial analysis pinpointed Tanimbar Selatan as a high-risk hotspot, while network analysis revealed a centralized inspection structure lacking horizontal connectivity between regions. The study's main contribution is recommending a spatial- and risk-based inspection strategy that is adaptive to the unique geographical challenges of archipelagic regions, emphasizing horizontal connectivity, sub-district clustering, and integrating GIS and social network analysis for adaptive, data-driven, and sustainable supervision.]]>
