Association, Reflection, Stimulation: Problem Exploration in Early Design through AI-Augmented Mind-Mapping

The formulation of a problem is often more essential than its solution, which may be merely a matter of mathematical or experimental skill. To raise new questions, new possibilities, to regard old problems from a new angle requires creative imagination and marks real advances in science. - Albert Einstein This dissertation aims at developing a computational framework to support the process of problem exploration in early design. To do so, we investigate digital mind-mapping as a tool for problem exploration and develop new algorithms and interaction workflows by leveraging large knowledge databases. The central premise of this work is that channeling the designer's thinking process through intelligent stimulation using such databases can augment designers' ability to reason about the problem at hand and creatively synthesize new ideas to address the problem. Design problems are typically ambiguous, ill-defined, unstructured, and open-ended. Therefore, learning about the problem and exploration of the problem domain is critical in early design to build a well-developed understanding of the context toward fruitful solution exploration in design. Despite the importance of problem understanding in design, little research has been devoted to investigating problem exploration activities in-depth and drawing a clear connection on the effects of such activities on the resulting design outcomes. Most current efforts focus exclusively on implementing methods for ideation, conceptualization, and concept evaluation wherein the solution space takes prominence. In this regard, this dissertation aims to complement this with a study of problem exploration techniques (mind-mapping and free writing) and evaluation in early design. We highlight the importance of problem-based exploration and learning, and share insights on how the structure and associative capability afforded by mind-maps affect ideation on the problem statement, product opportunity gap, and the needs around a given design context. It is common for designers to tend to commit to solutions too early and limit the potential of discovering creative and novel ideas in early design. This tendency is further pronounced with the advent of several digital design tools that are feature-rich and focus on design conceptualization and solution formulation, rather than design problem exploration. Additionally, much of the research in design theory and methodology has also mostly focused on conceptualization techniques such as C-Sketch and morphological matrix, that aim to support the formation of new solution concepts through modification and re-interpretation of rough initial ones. To complement these, in this dissertation, we emphasize the importance of problem exploration and brainstorming tasks towards design opportunity identification during early design. This is studied with the use of mind-maps, a technique that helps designers express their thoughts by making connections or associations between ideas around a given context. Further, we propose novel human-computer collaborative mind-mapping workflows for enhancing design experiences through novel textual, verbal and visual computer supports. Specifically, we designed and implemented two cognitive support mechanisms to help designers in inspecting design problems and generating ideas. Human-subject studies were conducted to examine how these systems perform and user perception. Based on the extensive investigation, this dissertation further shares insights on how to promote reflection in problem exploration by stimulating association across ideas, and develops design implications for intelligent AI-augmented workflows during early design exploratory tasks