Information visualization for informed decision making in design space exploration by shopping.

Summary In design space exploration, the data resulting from the simulation of a large number of design alternatives can lead to information overload when it comes to choosing a good design solution. This design space exploration is similar to a multi-criteria design optimization method but in a manual mode for which appropriate multi-dimensional data visualization tools are employed. For the designer, a three-phase process - discovery, optimization, selection - is followed in a Design by Shopping paradigm. By "shopping" the design space, intuition about feasible and infeasible solution subspaces and solutions with good tradeoffs is gained. The designer learns during these graphical data manipulations. The selection of an optimal solution is thus based on an informed decision. The objective of this research is the performance of graphical representations for design space exploration, for the three phases of the Design by Shopping process. For this purpose, five graphical representations, identified as potentially performing, are tested through two experiments. In the first one, thirty participants tested three graphs, for the selection phase in a multi-attribute situation, through three design scenarios where a car has to be chosen among forty according to stated preferences. For this, a quality index is proposed to calculate the quality of the designer's solution for one of the three defined scenarios, the optimal solution according to this index being compared to those obtained after manipulation of the graphs. In the second experiment, forty-two novice designers solved two design problems using three graphs. In this case, the performance of the graphs is tested for informed decision making and for the three phases of the process in a multi-objective situation. The results reveal that one graph is suitable for each of the three phases of Design by Shopping:: the Scatter Plot Matrix graph for the discovery phase and for informed decision making, the Simple Scatter graph for the optimization phase and the Parallel Coordinate Plot graph for the selection phase in both a multi-attribute and multi-objective situation.