The Role of Data Visualization Tools in Enhancing Decision-Making Quality During High-Stakes Public Service Operations

High-stakes public service operations such as emergency response coordination, disaster management, healthcare triage, and national security monitoring require rapid, data-driven decision-making under conditions of uncertainty, time pressure, and information overload. This paper presents a novel Adaptive Multi-Modal Visualization Optimization Framework (AMVOF) designed to enhance decision-making quality by dynamically integrating heterogeneous data streams into cognitively efficient visual representations. The framework leverages a hybrid architecture combining Graph Neural Networks (GNNs) for relational data structuring, Temporal Convolutional Networks (TCNs) for real-time trend extraction, and a newly proposed Cognitive Load-Aware Visualization Selection Algorithm (CLVSA) that adaptively selects optimal visualization formats (e.g., heatmaps, network graphs, temporal dashboards) based on operator context and task criticality.