View-Flattening: Revealing Heterogeneous Multi-Dimensional Data Attributes within a Single View

Journal Title
Journal ISSN
Volume Title
The drastic increase in data is impacting all knowledge workers and driving demand for effective methods of exploring, analyzing, and understanding data. Much of this ever-growing data is heterogeneous multi-dimensional data. In other words, it may have many multi-typed data attributes. Currently, the tendency is to approach this complexity through a type of divide and conquer technique by making multiple views using largely familiar visualizations, and providing types of filtering interactions. In this PhD research, we examine whether it is possible to try a different approach to this problem by considering to what extent existing or familiar visual representations can be extended to incorporating multiple heterogeneous data attributes into a single view. This is the idea behind a concept that we call view-flattening. To explore this idea, we have designed and implemented new interactive visualizations of heterogeneous multi-dimensional data in three domains including security, business intelligence, and health care. In developing these visualizations, we followed a traditional human-centered methodology, working closely with domain experts to understand their data needs and tasks. Designing different domain-specific flattened visualizations acted as a catalyst in this research, through which we derived a generalized view flattening technique that can extend several well-known and familiar visualizations. Our explorations with this technique have resulted in a design space of extended familiar visualizations that can offer expanded possibilities for embedding attributes within a visualization view.
Hosseinkhani Loorak, M. (2018). View-Flattening: Revealing Heterogeneous Multi-Dimensional Data Attributes within a Single View (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from doi:10.11575/PRISM/5459