Visual Analytics for Understanding Relationships between Entities

A graph is a mathematical model for representing a system of pairwise relationships between entities. The term “graph” or “graph data” is quite often used to refer, actually, to a system of relationships, which can be represented as a graph, rather than to the mathematical model itself. In line with this practice, the term “graph” is used in this chapter as a synonym to “system of relationships”. Graph data have high importance in many application domains, such as social network analysis, transport network analysis, systems biology, transaction analysis, to name a few. The relational nature of graph data sets presents unique challenges that open up space for distinctive and innovative visual analytics methods. Graphs often contain inherent hierarchies, and the analysis outcomes vary significantly depending on the scale (i.e., aggregation level) at which the relationships are considered. Identification, analysis, and interpretation of inherent structures, such as tightly connected parts of the network, requires a combination of algorithmic and interactive methods. Analysis of graph structure may require involvement of different graph theoretic properties of nodes and links, consideration of paths, and extraction of recurring connectivity patterns. All of these analyses are taken to a different level of complexity when these graphs are multivariate, i.e., where each node carries multiple attributes, and dynamic, i.e., when the structure of the graph changes with time. In this chapter, we discuss these challenges in depth and present examples of employing visual analytics techniques for extracting valuable findings from this highly rich and interesting data type.