Spatial-Temporal Visualization of Tuberculosis Vulnerability in Surabaya, Indonesia, Using K-Means Clustering

Tuberculosis (TB) is a highly infectious disease in Indonesia, including Surabaya. The high number of Tuberculosis cases in Surabaya highlights the need for vulnerability mapping to support prevention and treatment. This research aims to build a website-based spatial and temporal visualization system that can display the level of Tuberculosis vulnerability in each sub-district in the city of Surabaya. The data used included five criteria: the number of Tuberculosis cases, the number of substandard housing units, population density, the number of health facilities, and sanitation. Before clustering, a normalization process was carried out using the Simple Additive Weighting (SAW) method was applied to standardize each variable to a scale of $0-1$, ensuring fair contribution to clustering regardless of unit differences. K-Means Clustering was then used to group the data, and clusters were labeled based on the average total of all normalized variables. Clusters with higher averages were labeled as low vulnerability, followed by medium and high. The labeled clusters were visualized on an interactive map using open map data. The system successfully grouped regions into three clusters and visualized them annually from 2019 to 2023. Evaluation showed an average $\mathbf{R}^{\mathbf{2}}$ score of $\mathbf{0. 5 4 2}$, indicating that the visualization reflects real-world vulnerability levels across sub-districts. This system can be a valuable tool for the Surabaya City Health Office to identify priority areas for intervention, plan prevention programs, and support data-driven decision-making efficiently and accurately.