Design Perspectives in Visualising Complex Information

We use an increasing number and variety of systems to present information. Many of the early visualisation systems concentrated on scientific information where dimensions were low in number, objective and orthogonal e.g. in visualising fluid flow over a turbine, or population density over a nation. Recently, however, more complex information has been the subject of spatial models, for example databases of documents, patterns of economic and financial data, and design spaces for electronic circuits. Data dimensionality may be in the thousands, and each dimension may have complex interrelationships with many others. In designing systems for interaction with such data, a variety of topics become relevant which were mostly ignored in early systems. Following on from a brief overview of my own work and some of the lessons and problems arising from it, three points of view are offered—spatial, semantic and social— on the design of spatial information systems. It should be emphasised that these three do not partition the relevant issues. Just as in the more general topic of design, they weave through each other with many interconnections. By looking from each of these perspectives, it is suggested that we need to broaden the range of topics considered relevant to the design of information visualisation systems.