Invention arises from novel combinations of prior technologies. However, prior studies of creativity have suggested that overly novel combinations may be harmful to invention. Apart from the factors of expertise, market, etc., there may be such a thing as ‘too much’ or ‘too little’ novelty that will determine an invention’s future value, but little empirical evidence exists in the literature. Using technical patents as the proxy of inventions, our analysis of 3.9 million patents identifies a clear ‘sweet spot’ in which the mix of novel combinations of prior technologies favors an invention’s eventual success. Specifically, we found that the invention categories with the highest mean values and hit rates have moderate novelty in the center of their combination space and high novelty in the extreme of their combination space. Too much or too little central novelty suppresses the positive contribution of extreme novelty in the invention. Furthermore, the combination of scientific and broader knowledge beyond patentable technologies creates additional value for invention and enlarges the advantage of the novelty sweet spot. These findings may further enable data-driven methods both for assessing invention novelty and for profiling inventors, and may inspire a new strand of data-driven design research and practice.

Design innovation projects often generate large numbers of design ideas from designers, users, and, increasingly, the crowd over the Internet. Such idea data are often used for selection and implementation but, in fact, can 1also be used as sources of inspiration for further idea generation. In particular, the elementary concepts that underlie the original ideas can be recombined to generate new ideas. But it is not a trivial task to retrieve concepts from raw lists of ideas and data sources in a manner that can stimulate or generate new ideas. A significant difficulty lies in the fact that idea data are often expressed in unstructured natural languages. This paper develops a methodology that uses natural language processing to extract key words as elementary concepts embedded in massive idea descriptions and represents the elementary concept space in a core–periphery structure to direct the recombination of elementary concepts into new ideas. We apply the methodology to mine and represent the concept space underlying massive crowdsourced ideas and use it to generate new ideas for future transportation system designs in a real public sector-sponsored project via humans and automated computer programs. Our analysis of the human and computer recombination processes and outcomes sheds light on future research directions for artificial intelligence in design ideation.

In order to develop novel solutions for complex systems and in increasingly competitive markets, it may be advantageous to generate large numbers of design concepts and then to identify the most novel and valuable ideas. However, it can be difficult to process, review, and assess thousands of design concepts. Based on this need, we develop and demonstrate an automated method for design concept assessment. In the method, machine learning technologies are first applied to extract ontological data from design concepts. Then, a filtering strategy and quantitative metrics are introduced that enable creativity rating based on the ontological data. This method is tested empirically. Design concepts are crowd-generated for a variety of actual industry design problems/opportunities. Over 4000 design concepts were generated by humans for assessment. Empirical evaluation assesses: (1) correspondence of the automated ratings with human creativity ratings; (2) whether concepts selected using the method are highly scored by another set of crowd raters; and finally (3) if high scoring designs have a positive correlation or relationship to industrial technology development. The method provides a possible avenue to rate design concepts deterministically. A highlight is that a subset of designs selected automatically out of a large set of candidates was scored higher than a subset selected by humans when evaluated by a set of third-party raters. The results hint at bias in human design concept selection and encourage further study in this topic.

No abstract

Automation has enabled design of increasingly complex products, services, and systems. Advanced technology enables designers to automate repetitive tasks in earlier design phases, even high level conceptual ideation. One particularly repetitive task in ideation is to process the large concept sets that can be developed through crowdsourcing. This paper introduces a method for filtering, categorizing, and rating large sets of design concepts. It leverages unsupervised machine learning (ML) trained on open source databases. Input design concepts are written in natural language. The concepts are not pre-tagged, structured or processed in any way which requires human intervention. Nor does the approach require dedicated training on a sample set of designs. Concepts are assessed at the sentence level via a mixture of named entity tagging (keywords) through contextual sense recognition and topic tagging (sentence topic) through probabilistic mapping to a knowledge graph. The method also includes a filtering strategy, the introduction of two metrics, and a selection strategy for assessing design concepts. The metrics are analogous to the design creativity metrics novelty, level of detail, and a selection strategy. To test the method, four ideation cases were studied; over 4,000 concepts were generated and evaluated. Analyses include: asymptotic convergence analysis; a predictive industry case study; and a dominance test between several approaches to selection of high ranking concepts. Notably, in a series of binary comparisons between concepts that were selected from the entire set by a time limited human versus those with the highest ML metric scores, the ML selected concepts were dominant.