Interactive Details on Demand Visual Analysis on Large Attributed Networks

Increasing scale leaves a challenging problem for visualizing large attributed networks. This paper proposes a details on demand approach for exploratory visual analysis on large attributed networks. Major structures are located and emphasized at each level, providing clues for user observation. The detailed subnet structure emerges gradually through the exploration process. Our method dynamically aggregates network with consideration of both structural and attribute properties. It allows a flexible control of the hierarchy structure. A userspecified interaction strategy is introduced to enable users to customize the analysis flow according to different analytic tasks. Case studies demonstrate that the proposed method is effective in extracting global knowledge, locating major structures, and discovering hidden information in networks.