Biologically Meaningful Keywords for Functional Terms of the Functional Basis

AbstractBiology is recognized as an excellent source of analogies and stimuli for engineering design. Previous work focused on the systematic identification of relevant biological analogies by searching for instances of functional keywords in biological information in natural-language format. This past work revealed that engineering keywords could not always be used to identify the most relevant biological analogies as the vocabularies between biology and engineering are sufficiently distinct. Therefore, a retrieval algorithm was developed to identify potential biologically meaningful keywords that are more effective in searching biological text than corresponding engineering keywords. In our current work, we applied and refined the retrieval algorithm to translate functional terms of the functional basis into biologically meaningful keywords. The functional basis is widely accepted as a standardized representation of engineering product functionality. Therefore, our keywords could serve as a thesaurus for engineers to find biological analogies relevant to their design problems. We also describe specific semantic relationships that can be used to identify biologically meaningful keywords in excerpts describing biological phenomena. These semantic relations were applied as criteria to identify the most useful biologically meaningful keywords. Through a preliminary validation experiment, we observed that different translators were able to apply the criteria to identify biologically meaningful keywords with a high degree of agreement to those identified by the authors. In addition, we describe how fourth-year undergraduate mechanical engineering students used the biologically meaningful keywords to develop concepts for their design projects.PDF

Related Resources

See what’s new.



Towards a Digital Knowledge Base of Circular Design Examples through Product Teardowns

Leveraging teardowns, a commonly practiced activity among product…



PhenoLines: Phenotype Comparison Visualizations for Disease Subtyping via Topic Models

PhenoLines is a visual analysis tool for the interpretation of disease…



SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebooks

We present SkexGen, a novel autoregressive generative model for…



Robotic assembly of timber joints using reinforcement learning

In architectural construction, automated robotic assembly is…

Get in touch

Something pique your interest? Get in touch if you’d like to learn more about Autodesk Research, our projects, people, and potential collaboration opportunities.

Contact us