Generative Design (GD) produces many design alternatives and promises novel and performant solutions to architectural design problems. The success of GD rests on the ability to navigate the generated alternatives in a way that is unhindered by their number and in a manner that reflects design judgment, with its quantitative and qualitative dimensions. I address this challenge by critically analyzing the literature on design space navigation (DSN) tools through a set of iteratively developed lenses. The lenses are informed by domain experts' feedback and behavioural studies on design navigation under choice-overload conditions. The lessons from the analysis shaped DesignSense, which is a DSN tool that relies on visual analytics techniques for selecting, inspecting, clustering and grouping alternatives. Furthermore, I present case studies of navigating realistic GD datasets from architecture and game design. Finally, I conduct a formative focus group evaluation with design professionals that shows the tool's potential and highlights future directions.
Copyright is held by the author(s).
This thesis may be printed or downloaded for non-commercial research and scholarly purposes.
Supervisor or Senior Supervisor
Thesis advisor: Erhan, Halil
Member of collection