A0931
Title: An efficient method to scatter network nodes on a spherical surface via swarm intelligence
Authors: Chao-Hui Huang - National Tsing Hua University (Taiwan) [presenting]
Frederick Kin Hing Phoa - Academia Sinica (Taiwan)
Abstract: The space-filling problem has been an important topic in many scientific and practical aspects. There have been many theoretical and methodological results when the space is flat and regular, but few discussions are found for evenly distributing points on a spherical surface. Fibonacci lattice is an elegant solution to this problem when the points are all independent, but the condition is hardly fulfilled especially when we consider the nodes in a network. Although this problem easily becomes very complex and time-consuming with the existence of clusters, it is highly useful and practical in the visualization of network data. We provide an efficient two-steps method to arrange the network nodes uniformly on a spherical surface. We partition the spherical surface associated with a criterion about the edge/point ratio, then we scatter the nodes on the respective subspace according to the relationship between nodes and modularity. In order to reduce the computational efforts, we first uniformly distribute points on a two-dimensional plane uneven with a functional gradient, then we stereographically project all points from a gradient plane back to a sphere. Some networks are used for demonstration.
