Search for: All records

Abstract semantic attributes of designs (e.g., comfortable, luxurious, and durable) play a significant role in the assessment of user-facing products, capturing intangible factors that people may consider aside from performance requirements. However, due to the difficulty of mapping highly subjective and varying perceptions to specific design features, it remains a challenge to quickly and accurately translate these qualities into designs using computational design tools. Seeking to align computational and human representations of subjective design information, we investigate the utility of adapting representations of semantic attributes to designers’ perceptions through interactive models. A study is conducted in which users evaluate parameterized drinking mugs, indicating their perceptions of how comfortable each is to hold. Interactive Bayesian optimization is used to adaptively arrive at a design that optimizes this subjective quantity for each participant individually. Participants (N = 31) guide the model by providing their own decisions or building off of empirical data from a prior group of participants (N = 25). The resulting designs are evaluated across different scenarios, demonstrating the extent to which outputs of noninteractive models can be used to represent a subjective, semantic attribute and how interactive models may improve perceived alignment between human intent and computionally generated outputs. 

Abstract The development of example-based design support tools, such as those used for design-by-analogy, relies heavily on the computation of similarity between designs. Various vector- and graph-based similarity measures operationalize different principles to assess the similarity of designs. Despite the availability of various types of similarity measures and the widespread adoption of some, these measures have not been tested for cross-measure agreement, especially in a design context. In this paper, several vector- and graph-based similarity measures are tested across two datasets of functional models of products to explore the ways in which they find functionally similar designs. The results show that the network-based measures fundamentally operationalize functional similarity in a different way than vector-based measures. Based upon the findings, we recommend a graph-based similarity measure such as NetSimile in the early stages of design when divergence is desirable and a vector-based measure such as cosine similarity in a period of convergence, when the scope of the desired function implementation is clearer.