EAGER: Link Free Graph Visualization for Exploring Large Complex Graphs

点击翻译按钮获取中文摘要

This project develops proof-of-concept examples for a novel graph visualization technique named Link Free Graph Visualization (LFGV). The research addresses problems in existing graph visualization techniques: (1) becoming cluttered when visualizing large graphs; (2) having limited applicability to complex graphs, such as graphs with high dimensional node attributes and time evolving graphs; (3) misleading users due to the information loss caused by 2D or 3D node layouts and efforts toward increasing scalability. The research leads to new graph visualization techniques with better scalability, better applicability to complex graphs, and less misleading insights than NLDs. LFGV first projects a graph to a multidimensional space while preserving major graph topology information. LFGV then uses extended multidimensional visualization techniques to visualize the projection for exploring graph topology and other information carried by the graph, such as multidimensional node attributes. This project develops a working LFGV prototype for plain/multivariate graphs and conducts evaluations to prove that LFGV is viable. It also develops a weblog visual analysis approach to provide a test bed and an application example for LFGV. Graph visualization is widely used in lots of applications, such as social network analysis, bioinformatics, and web information exploration. LFGV impacts the application fields where large complex graphs need to be analyzed. The weblog visual analysis approach has a direct impact on online information browsing, retrieval, and analysis. The project publishes software for research and educational purposes.