Mapping the Colocalization Network: A Wayfinding Approach to Interacting with Complex Network Diagrams

Although network visualizations are becoming increasingly common, designing such visualizations can be challenging due to the number of visual elements and non-linear relations that they need to display. The main design challenge faced is finding the right trade-off between providing a sufficient level of information detail while keeping the visual complexity of the visualization as low as possible. One way of overcoming this challenge is to rely on the use of mental models that are familiar to the users of network visualizations. In this paper, we propose the use of a mental interaction model similar to that of map visualizations - generally based on geographical maps - as the basis for visual design of network diagrams. We argue that such a mental model would foster a set of network interaction tasks that can be defined broadly as wayfinding. We present the process of wayfinding from a semiotic standpoint, and match its main key points to those of interaction tasks with network diagrams. As a case study for this analysis, we also present a prototype network diagram visualization tool, called Colocalization Network Explorer, which we have developed to support the exploration of the relationships between various diseases and the portion of the human genome that is potentially involved in their onset. Additionally, we describe how the design process has benefited from the adoption of the wayfinding mental model.