BERT-based coupling evaluation of biological strategies in bio-inspired design. (15th July 2023)

• Record Type: Journal Article
• Authors: Sun, Feng; Xu, He; Meng, Yihan; Lu, Zhimao; Gong, Chengju
• Is Part Of: Expert systems with applications. Volume 222(2023)
• Publication Date: 2023-07-15
• Subjects: Bio-inspired design -- BERT -- Coupling evaluation -- Biological strategy; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2023.119725
• Languages: English
• ISSNs: 0957-4174

Abstract

Abstract: Searching for suitable biological strategies in bio-inspired design (BID) is the first problem that designers need to solve. Based on the biological strategy database of the AskNature, a natural language processing (NLP) method is applied to search for keywords related to design topics, and a multi-criteria coupled evaluation method is proposed for biological strategy. This paper firstly establishes the mapping relationship between BID topic and biological strategies through a NLP method-BERT. Then a coupled evaluation method is established based on the ordinal relation analysis and the deviation maximization method, highlighting the differences between biological strategies while considering the designer's preference. Finally, cases show the biological strategy coupling evaluation process. The results indicate that the proposed method is effective and efficient in biological strategy evaluation, which assists BID well. Highlights: BERT with coupling evaluation (CE) show efficiency in biological strategies selection. BERT with CE is presented with ordinal relation analysis and deviation maximization. Rank of the biological strategies indicate potential application in Biomimetic Design.