| |
|
FINDING DESIGN ANALOGIES
The search for ways to help engineering designers be more efficient, innovative, and inspired is ongoing, whether it is through trying to understand the design process from a cognitive perspective to better facilitate its success, or creating and testing new methods and tools for designers to use during the design process. Design by analogy, in which designers draw inspiration from cross-domain design solutions, is a promising methodology for industry use. My more recent work attempts to leverage the existing design solutions within a repository, combined with an exploration of inherent structural forms that can be discovered based on the content and similarity of that data, in order to gain useful insights into the nature of the design space. In this work, the approach will be applied to uncover structure in the U.S. patent database. This exploration of structures will be performed in a number of different ways, including examining different types of structures and their implications regarding the similarity among patents, and exploring function-based structures vs. surface-content-based structures, among others. These insights could generate fodder for stimulating design for engineering designers, which will be tested with a cognitive engineering design study.
More specifically, the approach involves processing patent text using Latent Semantic Analysis (LSA) [1], producing a cosine similarity matrix. This matrix is then input into an algorithm developed by Kemp and Tenenbaum [2] to discover structural form in data sets. The resulting structures are labeled with common terms using LSA analysis once more. These structures can then be used to understand the patent space and the interrelatedness of technologies within the patents and patent space, hopefully providing useful insights to designers for finding and applying analogies and producing innovative design solutions.
This work has been accepted to ASME IDETC 2011 for presentation. Fu, K., Cagan, J., Kotovsky, K., Wood, K., 2011, “Discovering Structure In Design Databases through Functional and Surface Based Mapping,” to be in the Proceedings of IDETC/CIE 2011, DETC2011-48322, Washington, DC.
- Landauer, T.K., Foltz, P.W. & Laham, D., 1998, “Introduction to Latent Semantic Analysis,” Discourse Processes, 25, 259-284.
- Kemp, C. and Tenenbaum, J., “The Discovery of Structural Form,” PNAS, 2008, Vol 105, no 31, 10687-10692.
Primary Researcher: KATE FU
|
|
