A computational framework for design-by-analogy using semantic-network representations
ABSTRACT: The framework proposed operationalized through four interrelated components: encoding, retrieval, mapping, and evaluation. Semantic networks serve as the underlying knowledge representation that enables information structuring and cross-domain association. By translating cognitive reasoning into a computational architecture, the proposed framework establishes a unified structure for supporting analogical design and provides theoretical and technical guidance for developing future semantic-network-based tools that more effectively facilitate creativity and innovation in conceptual design.
Authors
- Qianzhi Jing (ORCID: https://orcid.org/0000-0002-5025-6421)
- Haoyu Zuo (ORCID: https://orcid.org/0000-0003-3811-4479)
- Ruocong Liu
- Liuqing Chen (ORCID: https://orcid.org/0000-0002-2104-3697)
- Boheng Wang (ORCID: https://orcid.org/0000-0002-9875-6441)
- Kamyar Hazeri
- Zebin Cai (ORCID: https://orcid.org/0000-0002-2482-1765)
- Peter Childs
- Bowen Zhan (ORCID: https://orcid.org/0000-0001-8434-4705)
- Yuan Yin
- Saul Haydon Rowe
- Shafina Vohra
- Qingjue Wu
Institutions
- University of Wisconsin–Madison (US)
- Imperial College London (GB)
- Zhejiang University (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- Proceedings of the Design Society
- Published
- 2026-07-02
- DOI
- https://doi.org/10.1017/pds.2026.10403
- Primary Topic
- Design Education and Practice
- Type
- article
- Field-Weighted Citation Impact
- 0.00