Talk: Flow-based Influence Graph Visual Summarization

Speaker:
Prof. Lei Shi (http://lcs.ios.ac.cn/~shil/)
Institute of Software, Chinese Academy of Sciences

Time:
11am, August 4, 2014.

Bio:
Lei Shi is an associate research professor in the State Key Laboratory of Computer Science, Institute of Software, Chinese Academy of Sciences. Before 2012, he was a staff researcher, research staff member and research manager at IBM Research - China, working on information visualization and visual analytics. He holds B.S. (2003), M.S. (2006) and Ph.D. (2008) degrees from Department of Computer Science and Technology, Tsinghua University. His research interests span Information Visualization, Visual Analytics, Network Science and Networked Systems. He has published more than 50 papers in refereed conferences and journals. He is the recipient of IBM Research Accomplishment Award on "Visual Analytics" and the VAST Challenge Award twice in 2010 and 2012. He is now leading/participating a few national projects such as NSFC and the 973 program.

Abstract:
Visually mining a large influence graph is appealing yet challenging. People are amazed by pictures of newscasting graph on Twitter, engaged by hidden citation networks in academics, nevertheless often troubled by the unpleasant readability of the underlying visualization. Existing summarization methods enhance the graph visualization with blocked views, but have adverse effect on the latent influence structure. How can we visually summarize a large graph to maximize influence flows? In particular, how can we illustrate the impact of an individual node through the summarization? Can we maintain the appealing graph metaphor while preserving both the overall influence pattern and fine readability?

To answer these questions, we first formally define the influence graph summarization problem. Second, we propose an end-to-end framework to solve the new problem. Our method can not only highlight the flow-based influence patterns in the visual summarization, but also inherently support rich graph attributes. Last, we present a theoretic analysis and report our experiment results. Both evidences demonstrate that our framework can effectively approximate the proposed influence graph summarization objective while outperforming previous methods in a typical scenario of visually mining academic citation networks.

Joint work with Hanghang Tong, Jie Tang and Chuang Lin.