Multilevel compound tree - Construction visualization and interaction

Abstract

Several hierarchical clustering techniques have been proposed to visualize large graphs, but fewer solutions suggest a focus based approach. We propose a multilevel clustering technique that produces in linear time a contextual clustered view depending on a user-focus. We get a tree of clusters where each cluster - called meta-silhouette - is itself hierarchically clustered into an inclusion tree of silhouettes. Resulting Multilevel Silhouette Tree (MuSi-Tree) has a specific structure called multilevel compound tree. This work builds upon previous work on a compound tree structure called MO-Tree. The work presented in this paper is a major improvement over previous work by (1) defining multilevel compound tree as a more generic structure, (2) proposing original space-filling visualization techniques to display it, (3) defining relevant interaction model based on both focus changes and graph filtering techniques and (4) reporting from case studies in various fields: co-citation graphs, related-document graphs and social graphs. © IFIP International Federation for Information Processing 2005.