Fast Insights: Visualising Networks

Kimmo Soramäki and Samantha Cook

Network visualisations, where nodes are represented as circles or other shapes joined by links, are often the best way to understand a network’s structure. In addition to the basic structure of which nodes are linked to each other and, optionally, the direction of the links, network visualisations may show node and link properties, such as community classification or link weights. We have already seen many examples of network visualisations illustrating network structure and properties. This chapter will provide more detail on the mapping of node and link properties to visual elements – for example, node size or colour, and link width or opacity.

The most useful network visualisations are interactive, especially when networks are complex. For example, interactive visualisations may allow the user to zoom in and out, drag a node to change its position or select a node to show only it and its neighbours. Additional node and link information may be shown on hover, and advanced visualisations can allow the user to interactively change mappings and filter nodes or links. Such interactive visualisations conform to Schneiderman’s mantra of “overview first, zoom and filter, then details