Treemaps as a Tool for Social Network Analysis

摘要

We apply treemap technology in the analysis of large, social network datasets principally for examining network subgroups. A treemap is graphically- based information and exploration tool which is used in diverse fields such as computer science, finance, and human-gene research; from our experience, we find treemaps also useful in the social network analysis setting. A treemap represents hierarchical and categorical data in a mosaic form containing embedded, rectangular shapes, where the size of each shape is germane. Treemap displays are especially helpful when examining data in an interactive mode (as opposed to a static or printed form). We have found that treemaps are a powerful tool for exploring large social-networks, particularly during the exploratory data analysis phase. Their use quickly leads to a thorough perspective of the holistic characteristics of the network and to easier identification of significant subgroups; both of these perspectives may otherwise remain hidden using traditional visualization techniques. In this report, we introduce treemap technology, first broadly, then, specifically how it can be applied to social network analysis. We also show how we have actually applied treemaps to an interactive study of a large, real-world dataset. As a result of our experiences, we encourage other social network analysts to consider applying treemaps in their work.