Graph Analysis and Visualization: Discovering Business Opportunity in Linked Data

Chapter 6 Explore and Explain

After getting data, visually laying out the graph, and attaching data to visual attributes, you may want to explore the network in more detail to gain some insights. The interesting patterns described in the various examples in the previous chapters may not be so obvious in your graph. In many cases, some amount of interactive analysis is required before patterns emerge. Often, initial graphs look like “a hairball” or “a plate of spaghetti.” Do not despair. Most graph software packages contain interactive features to explore the graph in more detail, which are the topics of this chapter. For example, zoom and identification interactions let you explore graph details. Filters and selection help you focus on items of interest and hide less relevant items.

Assuming that you do gain some insights, in most situations you will want to share those findings with other people. Whether those findings appear in a PowerPoint presentation, whitepaper, or poster, you will want to help the viewer see those patterns, too. The second portion of this chapter covers enhancements such as annotations, labels, legends, and explanations to help you convey your findings.

Explore, Explain, and Export

Upon first viewing a graph, the viewer (whether the graph author or the presentation audience) may have some basic questions:

• What am I looking at?
• Are there some landmarks?
• What is this node (or link)?
• What is it connected to?

Basic interactions are used in ...