Towards a lexicon of visualization design templates: Supporting sensemaking with enhanced network visualizations

The information visualization community has grown expansively over the past few decades along with the digitization of information and the advances in computing technologies. As information availability continues to increase, visualization designers are challenged with the task of presenting massive volumes of data in a consumable manner. There has been minimal support for visualization designers in terms of guidance for developing effective visualizations based on an understanding of human perceptual and cognitive capabilities. This has largely been a result of the context-dependent nature of visualizations. Subsequently, this research effort provides an initial set of empirically-grounded design templates for network visualizations intended to support human sensemaking tasks. These templates are intended to be accessible to interface designers lacking formal training in human cognitive and perceptual capabilities. These templates represent a first step towards establishing a larger lexicon of design templates for multiple types of visualizations intended to support the design of visualizations that are grounded by empirical research to enhance the potential to support given audiences, tasks, and domains. To accomplish this goal, this research effort adopted two primary research aims. The first goal was to investigate empirically whether network visualizations are an effective and efficient means to support human sensemaking tasks compared to non-visualization displays. The second goal was to empirically demonstrate that network visualizations designed to support human behavior at different levels of cognitive control, by using the template-based designs, will better support sensemaking tasks than traditional network visualizations or non-visualization displays. Results showed that basic network visualizations often failed to offer any performance benefits over non-visualization displays for both information foraging and knowledge-based sensemaking tasks - frequently resulting in decreased performance, while the template-enhanced network visualizations provided significant performance benefits for both of these sensemaking tasks. Future work is needed to generate additional design templates for network and other visualization types and to support other common tasks; however, this research provides a necessary first step, including a framework for template creation and evaluation to guide future efforts.