A Dynamic Network Layout Visualization Method Based on Structural Similarity

摘要

Many real-world networks, including social and computer networks, have complex time-varying characteristics that evolve over time. Such dynamic networks are typically visualized using a sequence of timeslices, each timeslice presents the network topology at a particular time step. It is difficult to preserve the dynamic stability and visual continuity of the timeslices sequence, which allow researchers to interpret the temporal evolution of the network. In this paper, a dynamic network layout visualization method based on structural similarity is proposed to improve the quality of the timeslices. First, using network layout algorithm based on the community structure lay-outs a single timeslice and optimizing the layout through the static network cost function. Then, the new nodes are pre placed and the movement direction of nodes are constrained by a node movement method based on the s structural similarity, which can improve the dynamic stability and visual continuity of the timeslices sequence layout. Experiments show that the proposed method has a smaller layout cost in a single timeslice compared with the traditional dynamic network visualization, and a smaller node moving distance and the degree of structural shape change in the timeslices sequence.