Research

Large Displays and Visualization

Much of my research deals with visualizations on large displays. With large displays, it is possible to show one large visualization, or alternatively, many smaller visualization views. The figure illustrates the option of showing many smaller visualization views. I see this choice as principal, and as the most important consideration to keep in mind, when designing visualizations for large displays.

Displaying one large visualization gives room to subdivide the space into what we can consider as separate views. This option enables the designer to use spatial encoding to communicate the relations between individual views.

Displaying many smaller visualization views, designers can let users arrange these to make sense of data. This option leaves the task of communicating views’ relations to the designer, primarily using alternatives to spatial encoding.

Clearly, while these two options delimit the potential extremes of visualization views’ size, it is also possible to use them in combination. For example, by showing small views as overlays on top of a large visualization that fills an entire display.

In my work, I study both of these options. However, my focus has mainly been on using many small visualization views. This changes the manifestation of abundant display space. Here, abundant display space turns view-considerations into metaview considerations (i.e., considerations about or beyond the view), and changes the focus to letting users create new visualizations effectively and showing meta-visualizations of their relations.

Methodology

More to come…

Domain Studies

More to come…

Proxemics and Visualization

More to come…

Touch Interaction for Large Displays

More to come…

Multiple Views

The notion of Multiple views centers on the idea of showing data in separate views to allow people to see different parts of a data set separately, see the same data in different representations, or to let them lay out views manually to make sense of them, e.g., like people naturally do with documents on a table. The focus of my Marie Curie postdoctoral fellowship is to increase our understanding of how we can help people understand how views relate to each other. In our paper on multiple views, we coined the notion of view relation techniques. I study theoretical and formal aspects of multiple views, as well as very concrete uses rooted in this concept through collaborations with case partners in the health data domain. My case partners are concerned with analyzing complex data sets, where different disciplines need to work together to come to useful insights based on their data analysis process.

I consider large displays (see for example our paper on imaginative uses of large displays) as part of focusing on collaborative analysis processes, and more generally, the notion of ubiquitous analytics. To study this aspect, I am part of a newly established Center for Health Informatics at the Foothills campus of University of Calgary. The center is physically located at the Foothills Campus at University of Calgary and houses large displays and motion tracking technology. This allows us to study how large displays might be used as part of a ubiquitous analytics process, where researchers can enter the visualization room with their own device, move data and visualizations to a shared surface, and take them with them again when a collaborative analysis session is over. In this sense, the notion of multiple views works across devices and form factors.

I also consider theoretical aspects of multiple views in VIVIR. With onset from the notion of view relations, I build a formalization of the information visualization pipeline. This work will provide a strong formal understanding of differences and similarities between visualizations.

Featured publications

These are publications that I consider to be particularly significant in terms of showcasing my focus on multiple views, on working with concrete problems, and on considering methodology.