In the information society of today, datasets have become increasingly complex in the context of size, data dimensionality and time-dependency. In fact, the current complexity of data is staggering, and our ability to collect data is increasing at a faster rate that our ability to analyze it.

Information visualization, the representation of data graphically rather than textually, attempts to solve this problem by exploiting the high-bandwidth human perceptual and cognitive capabilities to detect patterns and draw interferences from visual form. Specifically designed visual metaphors define how abstract data values, which typically lack form in physical reality, are translated into easily understandable visual elements.

A dataset is in essence a large collection of textual and numerical variables, connected by a complex network of meaningful relationships. In an architectural context, one can consider data as a raw material, which lives in a 'virtual' reality. Similar to our physical world, this material requires designers to shape an inhabitable and enjoyable world in which people can become enlightened by streams of information.

Following Louis Henry Sullivan’s design concept of ‘form follows function’, one can thus imagine a similar concept of ‘form follows data’, or the ‘architectural’ use of data bits as virtual bricks. Good design becomes a process of exploring the most effective formalistic data-mapping principles, an investigation to aesthetically translate the inherent structures hidden inside the data itself.

Several software packages specialized in the creation of visual and interactive artifacts have become increasingly affordable, better community supported and easier to use. Built-in features such as “real-time execution” and “visual programming blocks” have introduced new perspectives on the process of computer programming, in so far that ‘’software engineering’ has become remarkably similar to ‘design sketching’. Developers can perform experiments on the fly, focusing on conceptual elements rather than configuration issues.

Meanwhile, many datasets have become increasingly accessible, relevant and even cool and fashionable. Originating from a wide set of sources, including online electronic social networks, non-governmental organizations, Right to Information Acts or unfortunate corporate leaks, they appeal to our human curiousness and desire to learn.

As a result, a parallel stream of software developers, consisting of young and self-motivated people, has recently started to experiment representing these complex datasets in intriguing ways. Independent from institutional or commercial pressure, their works demonstrate initiative, enthusiasm, interest and skill to tackle complex issues which were previously typically reserved for the visualization researcher.

The adopted approaches often stand out in terms of artistic approach and visual aesthetic quality. They demonstrate that information visualization can be enriched with the principles of creative design and art, to develop valuable data representations that also address the emotional experience of users, moving away from the typical focus on typical task effectiveness metrics. They show that visualization can be broadened to encompass the stimulation of alternative human senses, such as sound, smell, touch and taste, as high-bandwidth parallel streams of sensory experiences, provoking emotion and interpretation.

As the generation of data and access to information increasingly penetrate everyday life as always-on, always-there interfaces, novel representation approaches are required that take into account user interest, attention, enjoyment and curiosity. As electronic billboards become as large as building facades and displays as small as electronic jewelry, the field of visualization stands for an exciting era in which information access becomes redefined.

Andrew Vande Moere is lecturer and Assistant Professor at the Key Centre of Design Computing and Cognition, University of Sidney. He is the editor of Information Aesthetics (http://www.infosthetics.com), a widely read weblog about data visualization.