Encoding Design Process Using Interactive Data Visualization

The existing research on design space exploration favors the exploration of multiple parallel designs, however the act of exploring a design space is still to be integrated in the design of new digital media. We conducted an experiment to understand how designers navigate through large numbers of design alternatives generated from parametric models. We analyzed the data with a purpose-built visualization tool. We observed that participants changed the task environment and took design actions, frequently combining these into action combinations. Five tasks emerged from our analysis: Criteria Building, Criteria Testing, Criteria Applying, Reflection and (Re)Setting. From our analysis, we suggest several features for future systems for interacting with design alternatives.