Efficient Node Overlap Removal Using a Proximity Stress Model

Abstract

Abstract. When drawing graphs whose nodes contain text or graphics, the nontrivial node sizes must be taken into account, either as part of the initial layout or as a post-processing step. The core problem is to avoid overlaps while retaining the structural information inherent in a layout using little additional area. This paper presents a new node overlap removal algorithm that does well by these measures.

Node Overlap Removal for 1D Graph Layout

“…Existing techniques for removing node overlapping in 2D graphs (e.g., [14], [16]) move nodes from the initial embedding until no more overlapping exist in the layout (see Section VI for more details). As a result, distances between pairs of nodes in the final layout depend on how nodes were overlapping in the initial layout and the Requirement 4 is not always met.…”

“…This is due to the fact that it is the only one exclusively devoted to solve the 1D version of the problem. We did not include the approach proposed by [16] because its adaptation to the 1D case is not trivial. …”

“…When dealing with large graphs, an important drawback of MDS techniques is that it produces embeddings where many nodes overlap, leading to cluttering issues. Several algorithms have been proposed to deal with this issue for 2D graph layout [11], [12], [13], [14], [15], [16], [17]. The purpose of these works is to avoid overlapping and try to preserve the initial configuration of the nodes as much as possible.…”

Interactive Network Exploration to Derive Insights: Filtering, Clustering, Grouping, and Simplification

“…One approach is to count the number of flaws that inhibit readability [27] and support analysts in reducing their prevalence [2,7,12].…”

More Flexible Radial Layout

“…Quite recently, extensions of the stress term have been used for drawing graphs with explicitly formulated aesthetic criteria, such as the uniform scattering of the nodes in a graph over a unit disk [16], penalizing node overlaps [11], or preserving a given topology [10].…”