Visualizing Large Graph Database in Visual Data Management Systems (VDMS)
Jaideep Reddy Gade, Uzair Hamed Mohammed, Luis Aguilar
Department of Computer Science
Faculty Supervisor: Shahrukh Humayoun
This project enhances GraDVis, an interactive graph database visualization tool for Visual Data Management Systems (VDMS), by addressing its performance and interactivity limitations. The original system struggled to render graphs exceeding 10,000 nodes and lacked insightful visual exploration capabilities. To overcome these challenges, we developed a WebGL-based web application that significantly improves large-scale graph rendering and user interaction. The system integrates SigmaJS, React Sigma, and Graphology for efficient force-directed graph visualization, while D3.js enables an interactive tree visualization to illustrate customer journeys within retail datasets. The backend is implemented using NodeJS and Express to manage graph data and user requests. Performance was evaluated through controlled interaction experiments measuring average frames per second (FPS) across varying dataset sizes (100 to 29,000+ nodes) in both high- and low-performance computing environments. Results demonstrate that the WebGL-based implementation substantially improves scalability and maintains interactive performance above acceptable thresholds, offering a more robust and insightful tool for exploring large graph databases.