Design-by-Analogy: Exploring for Analogical Inspiration With Behavior, Material, and Component-Based Structural Representation of Patent Databases

Abstract

Design-by-analogy (DbA) is an important method for innovation that has gained much attention due to its history of leading to successful and novel design solutions. The method uses a repository of existing design solutions where designers can recognize and retrieve analogical inspirations. Yet, exploring for analogical inspiration has been a laborious task for designers. This work presents a computational methodology that is driven by a topic modeling technique called non-negative matrix factorization (NMF). NMF is widely used in the text mining field for its ability to discover topics within documents based on their semantic content. In the proposed methodology, NMF is performed iteratively to build hierarchical repositories of design solutions, with which designers can explore clusters of analogical stimuli. This methodology has been applied to a repository of mechanical design-related patents, processed to contain only component-, behavior-, or material-based content to test if unique and valuable attribute-based analogical inspiration can be discovered from the different representations of patent data. The hierarchical repositories have been visualized, and a case study has been conducted to test the effectiveness of the analogical retrieval process of the proposed methodology. Overall, this paper demonstrates that the exploration-based computational methodology may provide designers an enhanced control over design repositories to retrieve analogical inspiration for DbA practice.

An exploration-based approach to computationally supported design-by-analogy using D3

“…Topic modeling is implemented similar to Divide-and-Conquer NMF (DC-NMF) (Du et al ., 2017). In the authors’ prior work, the topic modeling method was successfully implemented to cluster patent data by their semantic topics (Song and Fu, 2019). It has an advantage that it can efficiently process a large volume of data and it does not rely on the selection of an appropriate topic number k, which is considered a computational challenge when applying flat NMF on a given corpus.…”

“…The multiple representations facilitate the patent retrieval process, as the designer explores the patent database along the proposed analogical properties. Various design insights can arise from the unique representations of the design repository (Song and Fu, 2019). This procedure is achieved by processing the word-document matrix of patent data.…”

“…For instance, an engineering principle-based patent network is created by clustering patent data that are processed to contain only engineering terms [extracted from a glossary of engineering (Hanifan, 2014)]. For more detail about how different structural representations of patent data were created, see Song and Fu (2019). Filtering patents by label terms and CPC sections : The large dataset size and several edge crossings (geometric intersections of two or more edges) of the network can impose a challenge to a designer navigating around the spatial data.…”

“…), and material content (What materials does the system use or consume?) and evaluated their effect on the design process (Song and Fu, 2019). These findings demonstrate that unique design insights can arise from different representations of a patent dataset.…”

Invention engineering – a novel approach to support creativity in engineering, practice and research

“…This challenge comprises the development of techniques and technologies for use in areas outside engineering such as literature, music, visual arts, film et al To be able to discover ideas and concepts that can provide creative impulses for engineering research, IT technologies such as text mining, web crawling or machine learning are available [21,22]. Respective search results need to be referenced with a semantic model and processed in a knowledge system (Idea Bank), which -as a kind of associative search engine -provides input for idea generation.…”