Edge Bundling Techniques Toward Scalable Visualization of Large Graphs
Tuesday, September 5, 2017
4 p.m., Avery 115
3:30 p.m., Avery 347
Hongfeng Yu, Ph.D.Assistant Professor, Department of Computer Science and Engineering, University of Nebraska-Lincoln
Graphs, also called networks, are widely used to model relationships between data entities in diverse scientific and engineering applications, such as systems biology, software engineering, social science, and so on. Directly visualizing a large graph as a node-link diagram often incurs visual clutter. Edge bundling can effectively address this issue and concisely reveal the main graph structure with reduced visual clutter. Although researchers have devoted noticeable efforts to develop acceleration methods, it remains a challenging task to efficiently conduct edge bundling, particularly, on devices with limited computing capacity, such as ubiquitous smart mobile devices. In this talk I will briefly revisit existing edge bundling work, and introduce a few techniques developed in our recent work leading to performance optimization of edge bundling. I will also demonstrate a few applications enabled by these techniques and discuss possible future directions.
Hongfeng Yu is an assistant professor in Computer Science and Engineering at the University of Nebraska-Lincoln. He received his B.S. degree and M.S. degree in Computer Science from Zhejiang University, and Ph.D. degree in Computer Science from the University of California-Davis. After his Ph.D. graduation, he spent four years as a postdoctoral researcher at Sandia National Laboratories, California. His research concentrates on big data analysis and visualization, high-performance computing, and user interfaces and interaction. The research effort of his team has been leading to several new scalable algorithms and systems by exploiting high-performance parallel supercomputers and commodity programmable graphics hardware. These technologies have helped the scientists from several universities and national laboratories to obtain scalable visualizations for tera- and peta-scale applications, such as turbulent combustion, earthquake simulations, supernova evolution, climate simulations, and so on.