A design space of semantically-resonant pattern

Abstract

Semantic association is a concept from psychology which means two concepts can be connected together through the same experience [19]. In the field of information visualization, some studies have established such connections by selecting colors for data in charts [15, 22]. However, patterns, offering an effective alternative when colors may not be applicable, lack research and design experience related to their semantic association. In this thesis, I have studied effective methods to design semantically-resonant patterns and summarized a design space for others to reference.

In this thesis, I first presented a design workshop for designing semantically-resonant patterns. Specifically, I invited 13 participants with experience in design and information visualization. I collected 39 design sketches and generated 301 initial qualitative codings. After revising and refining these codings, I simplified them to 225 entries, which I then summarized in an affinity diagram for design space analysis.

I outlined the two steps for designing semantically-resonant patterns and defined the three methods for increasing concreteness of concepts used in step 1: the conceptual bridging method, the attribute extraction method, and the quantification method. I also categorized the visual variables used in step 2. I also categorized the visual variables used in step 2. Furthermore, I analyzed the design sketches to extract several key design insights and identified applicable scenarios for the three methods. Finally, I explored commonalities in pattern design and the potential for developing more complex semantically-resonant patterns.