The Physical Visualization of Information: Designing Data Sculptures in an Educational Context
Visual Information Communication
This paper is a qualitative case study analysis of the issues involved in designing and implementing data sculptures, the physical "embodiment" of data in a tangible presence, shape or form, within the context of an undergraduate design studio. It demonstrates how approaches and theories from the field of data visualization can form a rich contextual resource and motivational tool for teaching principles of creative design. A relatively short, two-week assignment challenged students to
... te and convey a meaningful data-driven insight through a tangible representation. In this paper, we analyze the resulting collection of physical artifacts developed by the students to reveal notable design approaches of embodying abstract information. We present a novel classification of data sculptures based on a semiotic taxonomy (symbolic, iconic, or indexical) and consider the qualities of representational fidelity and narrative formulation, for instance when the unfolding of the data-driven narrative is seen as a process, rather than an outcome. In addition, we investigate how the introduction of digital fabrication techniques influenced the design strategies chosen by the students versus handmade approaches. Information visualization (infovis) is traditionally viewed as a means to understand and analyze complex data sets. Its main tasks typically include: the accurate detection of data patterns, the facilitation of decision making based on data trends, and the communication of data-driven information and knowledge. With of its deep roots in scientific reasoning, research into information visualization has mainly focused on supporting scientific goals and optimizing analytic tasks. As a result, visualization is traditionally considered a tool that is neutral, dispassionate and objective, as it caters for the solutions to well-defined and specialized task of its expert users. In recent years, there has been a tendency to "democratize" the practice of data visualization, which is characterized by (1) the ability to develop visualizations through free, open and well-supported software development platforms by computer novices, (2) the increasing number of freely available, large, complex and interesting data sets from government, science and business sources, and (3) the refocusing of traditional data visualization goals towards including the education of a large, lay audience about data-supported insights. In the context of this last point, the fields of arts, design, media and sciences have a common interest in utilizing innovative forms of visualization to represent the knowledge, relevance and meaning hidden within complex information structures. By articulating the common interests and goals of visualization across these academic fields, shared problems are addressed in a multiperceptual space and a broader cultural context [1, 2, 3]. Several new visualization taxonomies have been proposed that incorporate the increasing artistic tendencies in the practice of data visualization. For instance, Warren Sack  has noticed how the "anti-sublime" information visualization has outgrown science and engineering and has ultimately become embraced by conceptual art. He proposes looking beyond whether such visualizations are pretty, beautiful, or easy to use, and instead recognize their aesthetics of governance -the interpretation and articulation of meaning, rather than just the purity of the information being represented, as a creative response to visual forms of contemporary art. Pousman et al.  coined the term casual visualization as conveying an "increased focus on activities that are less task driven, data sets that are personally meaningful, and built for a wider set of audiences". Viegas and Wattenberg  use the term artistic data visualization for those data depictions that embody a forceful point of view by recognizing the power of visualization as a potential communication medium. In turn, we defined information aesthetic visualization to denote the interlocking combination of aesthetics and design (e.g. visual style, user experience), a shift in focus, away from conveying patterns and trends towards narrating meaning and underlying principles, and finally, a tendency towards more fluid user interaction and user participation (e.g. flow, user feedback)  . The term information "aesthetics" here refers to the degree of involvement and engagement it facilitates, rather than denote a subjective measure for visual appeal or even the quality of the work. Similarly, the term database aesthetics has been used in the field of Design and Media Arts to denote a new aesthetic approach undertaken by a collection of artists who use the vast amounts of available information as their medium of expression   . At this extreme of what some might doubt can still be called "visualization", the visual ordering and organization of information has become an artistic choice, a choice motivated by the intent to influence and critique the digitization of daily life. This rich spectrum of approaches shows the importance of design and creativity in the visualization field. To explore the potential of design and of data visualization as a communication medium in its own right, the main author has coordinated and taught an undergraduate Design Studio at the University of Sydney. In particular, the studio course aimed to combine the fields of design, new forms of digital technology and data visualization as a platform for research-led teaching in design education. As such, one of its assignments comprised of the design and the implementation of a "data sculpture". The main aim of this paper is to report on the design strategies and experimentation by the students within the context of this assignment. It will present a novel classification of data sculptures based on a semiotic taxonomy (symbolic, iconic, or indexical) and the qualities of representational fidelity and narrative formulation.