Computational Support for Bridging Analogies

Abstract

Analogies are comparisons between two topics in terms of relational similarity, such as comparing a spring and a flexible table in terms of how they both exert explain upward force on your hand. Far domain analogies --- analogical comparisons between topics that seem very different on the surface, such as the solar system and an atom --- have been identified as being useful for creative ideation. However, people struggle to benefit from them. In this thesis, I explore how bridging analogies, analogies that bridge between a knowledge anchor that is familiar to the problem solver and a target analogy, can aid in allowing innovators to benefit from far domain analogies. Utilizing a breadth-first search in a graph of concepts from Wikipedia, we identified bridging analogies that connect a participant's knowledge anchor to a far domain analogy. We conducted a think-aloud study in which participants were asked to brainstorm on three design challenges, alternating whether they were provided only distant analogies or far analogies and bridging analogies tailored to their knowledge anchors. Using qualitative analysis of the think-aloud data we observed that bridging analogies aided participants in producing more abstract solutions instead of more direct translation of the analogies in their solution. Our results imply that bridging analogies can effectively aid innovators in benefitting from far domain analogies when creative problem solving.