DATA-DRIVEN DESIGN-BY-ANALOGY: Current Status and Prospects

Abstract

Design-by-Analogy (DbA) is a design approach that helps designers reduce design fixation and improve 
design ideation results by generating new possibilities, solutions, or concepts in a target domain based on 
inspiration from a source domain. New prospects for creating data-driven techniques and tools for DbA 
assistance have recently been made possible by the expanding availability of design databases as well as 
the quickly evolving data science and artificial intelligence technology. In this work, we review previous 
data-driven DbA research and group individual studies into four categories—analogy encoding, retrieval, 
mapping, and evaluation—based on the data, techniques, and applications. This article clarifies the current 
status of data-driven DbA research and compares it with the cutting edge of data science and AI research 
to highlight intriguing research possibilities and directions for the field. It is based on both structured 
analysis and detailed organic evaluation. Lastly, we suggest a hypothetical data-driven DbA system for 
the future that incorporates all of the principles.