文章学习（二）综述类：数据驱动的类比设计：最先进的技术和未来的方向

DATA-DRIVEN DESIGN-BY-ANALOGY:STATE OF THE ART AND FUTURE DIRECTIONS

摘要

  类比设计（DbA）是一种设计方法，指的是从源领域得到灵感产生新的思路、机会或者设计应用于目标领域。类比设计有益于改善设计师的设计理念和打破定式思维。最近，日益普及的设计数据库以及飞速发展的数据科学和人工智能技术为开发数据驱动的方法和支持DbA的工具提供了新的机会。在这项研究中，调查了现有的数据驱动的DbA研究，并且根据数据、方法、和应用分为四类：即类比编码、检索、映射和评估。本文基于细微有源头的综述和结构化分析，阐述了迄今为止数据驱动具有先进水平的DbA 研究，并将其与数据科学和人工智能研究的前沿进行基准测试，以确定该领域有前途的研究机会和方向。最后，提出了一个未来的概念数据驱动的DbA系统。

1.介绍

  DbA 是工程设计领域的一个热门研究领域，有支持DbA的各种方法和工具：

类比设计：对WordTree问题重新表示方法的研究
J S Linsey, A B Markman, and K L Wood. Design by analogy: A study of the WordTree method for problem re-representation. ASME Journal of Mechanical Design, 134(4):041009, 2012.
基于功能的类比设计：类比搜索的函数向量化方法
Jeremy Murphy, Katherine Fu, Kevin Otto, Maria Yang, Dan Jensen, and Kristin Wood. Function based design by-analogy: a functional vector approach to analogical search. ASME Journal of Mechanical Design, 136(10):101102, 2014.
类比设计：探索专利数据库的行为、材料和基于组件的结构表示的类比灵感
Hyeonik Song and Katherine Fu. Design-by-Analogy: Exploring for Analogical Inspiration With Behavior,Material, and Component-Based Structural Representation of Patent Databases. ASME Journal of Computing and Information Science in Engineering, 19(2):021014, 2019.
基于探索的方法以计算支持：使用D3类比设计
Hyeonik Song, Jacob Evans, and Katherine Fu. An exploration-based approach to computationally supported：design-by-analogy using D3. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 34(4):444–457, 2020.
类比思维：设计背景下的介绍
Ashok K Goel and L H Shu. Analogical thinking: An introduction in the context of design. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 29(2, SI):133–134, 2015.

  近年来，越来越多的设计数据库和快速发展的数据科学和人工智能技术为支持 DbA 的新数据驱动方法和工具提供了支持。例如，深度学习、知识图、自然语言理解和计算机视觉可能支持类比表示、检索、映射和评估过程 。

使用卷积神经网络推导专利图像的设计特征向量
Shuo Jiang, Jianxi Luo, Guillermo Ruiz-pava, Jie Hu, and Christopher L Magee. Deriving design feature vectors for patent images using convolutional neural networks. ASME Journal of Mechanical Design, 143(6):061405,2021.
使用技术语义网络的设计知识表示
Serhad Sarica and Jianxi Luo. Design Knowledge Representation with Technology Semantic Network. In Proceedings of the Design Society: International Conference on Engineering Design (ICED), Gothenburg,Sweden, Aug. 16-20, 2021.
通过知识距离指导数据驱动的设计构思
Jianxi Luo, Serhad Sarica, and Kristin L. Wood. Guiding Data-Driven Design Ideation by Knowledge Distance.Knowledge-Based Systems, 218:106873, 2021.
用外行的话来说： 从科学文本中提取半开放关系
Ruben Kruiper, Julian Vincent, Jessica Chen-Burger, Marc Desmulliez, and Ioannis Konstas. In Layman’s Terms:Semi-Open Relation Extraction from Scientific Texts. In Annual Meeting of the Association for Computational Linguistics (ACL), pages 1489–1500, 2020.
组合器：基于计算机的模拟方法创意生成工具
Ji Han, Feng Shi, Liuqing Chen, and Peter R N Childs. The Combinator: a computer-based tool for creative idea generation based on a simulation approach. Design Science, 4(e11):1–34, 2018.
基于类比推理和本体论的创造性想法生成的计算工具
Ji Han, Feng Shi, Liuqing Chen, and Peter R N Childs. A computational tool for creative idea generation based on analogical reasoning and ontology. Artificial Intelligence for Engineering Design, Analysis and Manufacturing,32(4):462–477, 2018.
一个基于数据驱动的创新概念生成和评估方法
Ji Han, Hannah Forbes, Feng Shi, Jia Hao, and Dirk Schaefer. A Data-Driven Approach for Creative Concept Generation and Evaluation. In Proceedings of the Design Society: DESIGN Conference, volume 1, pages 167–176, Online, 2020.

2.背景

  认知心理学表明，人们往往倾向于通过从已知问题中映射解决方案来解决给定的问题，即类比推理。类比推理过程的三阶段模型，包括检索、映射和评估。在检索和映射过程中，通常会利用三种类型的相似性，即字面、属性和关系相似性。在工程设计领域，使用类比推理生成新的设计解决方案或概念称为类比设计。研究表明，设计师利用至少六个不同的维度来比较目标产品和源产品：工作原理、结构、与人交互、功能、能量流动和物质流动。

  类比距离的研究有：

"近"和"远"的含义：结构设计数据库的影响和类比距离对设计输出的影响
Katherine Fu, Joel Chan, Jonathan Cagan, Kenneth Kotovsky, Christian Schunn, and Kristin Wood. The meaning of “near” and “far”: the impact of structuring design databases and the effect of distance of analogy on design output. ASME Journal of Mechanical Design, 135(2):021007, 2013.
类比距离会影响构思的表现吗？
V Srinivasan, Binyang Song, Jianxi Luo, Karupppasamy Subburaj, Mohan Rajesh Elara, Lucienne Blessing, and Kristin Wood. Does analogical distance affect performance of ideation? ASME Journal of Mechanical Design,140(7):071101, 2018.
专利刺激搜索及其对理念结果的影响
Binyang Song, V Srinivasan, and Jianxi Luo. Patent stimuli search and its influence on ideation outcomes.Design Science, 3(e25):1–25, 2017.
关于创新设计类比的好处和陷阱：基于类比距离、通用性和示例模式的构想性能
Joel Chan, Katherine Fu, Christian Schunn, Jonathan Cagan, Kristin Wood, and Kenneth Kotovsky. On the benefits and pitfalls of analogies for innovative design: Ideation performance based on analogical distance,commonness, and modality of examples. ASME Journal of mechanical design, 133(8):081004, 2011.
研究与此相反的语义刺激对设计概念创造力的影响
Ivey Chiu and Li H Shu. Investigating effects of oppositely related semantic stimuli on design concept creativity.Journal of Engineering Design, 23(4):271–296, 2012.
刺激模式的影响和个人创造力支持系统中的关联距离
Ross A Malaga. The effect of stimulus modes and associative distance in individual creativity support systems.Decision Support Systems, 29(2):125–141, 2000.
创意模仿：探索跨行业创新案例
Ellen Enkel and Oliver Gassmann. Creative imitation: exploring the case of cross-industry innovation. R&D Management, 40(3):256–270, 2010.

  DbA 功效自然受到设计师知识和记忆的限制，无法对设计问题进行潜在检索和映射，这表明了利用外部知识数据库作为来源和使用智能算法进行类比推理的基本价值。

3.研究方法

  最近提出的数据驱动的DbA工具：

Idea-Inspire 4.0：L Siddharth and Amaresh Chakrabarti. Evaluating the impact of Idea-Inspire 4.0 on analogical transfer of concepts. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 32(4):431–448, 2018.
TechNet ：Serhad Sarica, Jianxi Luo, and Kristin L Wood. TechNet: Technology semantic network based on patent data.Expert Systems with Applications, 142:112995, 2020.
B-link：Feng Shi, Liuqing Chen, Ji Han, and Peter Childs. A data-driven text mining and semantic network analysis for design information retrieval. ASME Journal of Mechanical Design, 139(11):111402, 2017.
一个无人监督的深度学习模型，Sketch-pix2seq，该模型使用重建和聚类损失进行训练，以便提取Quickdraw数据库中草图的基本形状特征
Zijian Zhang and Yan Jin. An Unsupervised Deep Learning Model to Discover Visual Similarity Between Sketches for Visual Analogy Support. In Proceedings of the 2020 ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE), page V008T08A003, Virtual, Online, Aug. 17-19, 2020.

4.数据驱动的 DbA 方法和工具：综述

  随着过去二十年数据科学的兴起，开发了各种数据驱动的方法和工具来支持 DbA。

根据预先定义的大型语义网络 （WordNet） 提出了 WordTree 方法，在语言上重新表示设计问题：
J S Linsey, A B Markman, and K L Wood. Design by analogy: A study of the WordTree method for problem re-representation. ASME Journal of Mechanical Design, 134(4):041009, 2012.
提出了一种功能矢量方法，根据功能基础模型中功能动词的字频率将设计文档表示为向量：
Jeremy Murphy, Katherine Fu, Kevin Otto, Maria Yang, Dan Jensen, and Kristin Wood. Function based design by analogy: a functional vector approach to analogical search. ASME Journal of Mechanical Design, 136(10):
101102, 2014.
功能上新颖的概念可以使用功能类比搜索方法生成：
Katherine Fu, Jeremy Murphy, Maria Yang, Kevin Otto, Dan Jensen, and Kristin Wood. Design-by-analogy:experimental evaluation of a functional analogy search methodology for concept generation improvement.Research in Engineering Design, 26(1):77–95, 2015.
一种类比检索方法，通过文本挖掘和基于RNN的模型用于从文本中提取关系结构：
Roozbeh Sanaei, Wei Lu, Lucinne T.M. Blessing, Kevin N. Otto, and Kristin L. Wood. Analogy retrieval through textual inference. In Proceedings of the 2017 ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE), page V02AT03A007, Cleveland,USA, Aug. 6-9, 2017.
一个类比搜索引擎，利用信息提取策略和双向 RNN 深度学习模型，搜索遥远但相关的设计刺激，以满足特定需求：
Karni Gilon, Joel Chan, Felicia Y Ng, Hila Liifshitz-Assaf, Aniket Kittur, and Dafna Shahaf. Analogy mining for specific design needs. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems,Montréal, QC, Canada, Apr. 21-26, 2018.
一个名为"解决"的混合倡议系统，以识别科学论文中的类别：
Joel Chan, Joseph Chee Chang, Tom Hope, Dafna Shahaf, and Aniket Kittur. SOLVENT : A Mixed Initiative System for Finding Analogies between Research Papers. In Proceedings of the ACM on Human-Computer Interaction, pages 1–21, New York, USA, 2018.
一系列基于计算机的工具来支持 DbA，包括检索器和组合器：
Ji Han, Feng Shi, Liuqing Chen, and Peter R N Childs. The Combinator: a computer-based tool for creative idea generation based on a simulation approach. Design Science, 4(e11):1–34, 2018.
Ji Han, Feng Shi, Liuqing Chen, and Peter R N Childs. A computational tool for creative idea generation based on analogical reasoning and ontology. Artificial Intelligence for Engineering Design, Analysis and Manufacturing,32(4):462–477, 2018.
利用专利数据库进行DbA探索：
DbA 支持系统通过将设计问题的描述重新描述为动词及其同义词来识别专利数据库中的适应性类比：
Anthony Mccaffrey and West Brookfield. Analogy Finder, 2016, US Patent.
InnoGPS，这是一种基于整个美国专利商标局 （USPTO） 数据库中构建的总技术空间的计算机辅助构想工具：
Jianxi Luo, Serhad Sarica, and Kristin L. Wood. Guiding Data-Driven Design Ideation by Knowledge Distance.
Knowledge-Based Systems, 218:106873, 2021.
从现有的生物灵感设计抽象数据库和产品结构表示数据库中提炼出生物适应及其相关产品结构:
Devesh Bhasin, Daniel A McAdams, and Astrid Layton. A product architecture-based tool for bioinspired function-sharing. ASME Journal of Mechanical Design, 143(8):814101, 2021.

5.DbA领域结构化分析

  下图为数据、应用、方法关联的总体情况：

5.1数据

  下图总结了 DbA 中使用的数据集，包括专门为 DbA 研究开发的数据库。

  可用于类比设计2D和3D的图像数据库：

Shape-Net：Angel X Chang, Thomas Funkhouser, Leonidas Guibas, Pat Hanrahan, Qixing Huang, Zimo Li, Silvio Savarese,Manolis Savva, Shuran Song, Hao Su, and Others. Shapenet: An information-rich 3d model repository.2015
the IKEA design dataset：Joseph J Lim, Hamed Pirsiavash, and Antonio Torralba. Parsing Ikea Objects: Fine Pose Estimation. In IEEEInternational Conference on Computer Vision (ICCV), pages 2992–2999, Sydney, NSW, Australia, Dec.1-8,2013.
mechanical CAD model dataset：Sebastian Koch, Albert Matveev, Zhongshi Jiang, Francis Williams, Alexey Artemov, Evgeny Burnaev, MarcAlexa, Denis Zorin, and Daniele Panozzo. Abc: A big cad model dataset for geometric deep learning. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 9601–9611, Long Beach, CA, USA,June 16-20, 2019.
FOBIE dataset（旨在支持自然启发工程的半开放关系提取） ：Ruben Kruiper, Julian Vincent, Jessica Chen-Burger, Marc Desmulliez, and Ioannis Konstas. In Layman’s Terms:Semi-Open Relation Extraction from Scientific Texts. In Annual Meeting of the Association for Computational Linguistics (ACL), pages 1489–1500, 2020.

5.2应用（略）

5.3方法

  DbA 研究中已经使用的基于 AI 的技术在下图中进行了总结。

深度学习方法解决设计问题：
CNN:
Shuo Jiang, Jianxi Luo, Guillermo Ruiz-pava, Jie Hu, and Christopher L Magee. Deriving design feature vectors for patent images using convolutional neural networks. ASME Journal of Mechanical Design, 143(6):061405,2021.
Zijian Zhang and Yan Jin. An Unsupervised Deep Learning Model to Discover Visual Similarity Between Sketches for Visual Analogy Support. In Proceedings of the 2020 ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE), page V008T08A003, Virtual, Online, Aug. 17-19, 2020.
RNN:
Serhad Sarica, Jianxi Luo, and Kristin L Wood. TechNet: Technology semantic network based on patent data.Expert Systems with Applications, 142:112995, 2020.
LSTM:
Roozbeh Sanaei, Wei Lu, Lucinne T.M. Blessing, Kevin N. Otto, and Kristin L. Wood. Analogy retrieval through textual inference. In Proceedings of the 2017 ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE), page V02AT03A007, Cleveland,USA, Aug. 6-9, 2017.
以下文章利用AI技术对视觉和语义类比问题解决进行了广泛的探索：
Gaetano Rossiello, Alfio Gliozzo, Robert Farrell, Nicolas Fauceglia, and Michael Glass. Learning relational representations by analogy using hierarchical siamese networks. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), volume 1, pages 3235–3245, Minneapolis, MN, USA, June 2-7, 2019.
Fereshteh Sadeghi, C. Lawrence Zitnick, and Ali Farhadi. Deep Visual Analogy-Making. In Proceedings of the 29th Conference on Neural Information Processing Systems (NIPS), pages 1882–1890, Montréal, QC, Canada,Dec. 7-12, 2015.
Hongjing Lu, Ying Nian Wu, and Keith J Holyoak. Emergence of analogy from relation learning. Proceedings of the National Academy of Sciences, 116(10):4176–4181, 2019.
Fereshteh Sadeghi, C Lawrence Zitnick, and Ali Farhadi. VISALOGY: Answering visual analogy questions.In Proceedings of the 29th Conference on Neural Information Processing Systems (NIPS), pages 1882–1890,Montréal, QC, Canada, Dec. 7-12, 2015.
Jing Liao, Yuan Yao, Lu Yuan, Gang Hua, and Sing Bing Kang. Visual attribute transfer through deep image analogy. ACM Transactions on Graphics, 36(4):1–15, 2017.
研究为类比设计任务开发可解释性AI模型：
Alejandro Barredo Arrieta, Natalia Díaz-Rodríguezb, Javier Del Ser, Adrien Bennetot, Siham Tabik, Alberto Barbado, Salvador Garcia, Sergio Gil-López, Daniel Molina, Richard Benjamins, and Others. Explainable Artifi-cial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI. Information Fusion, 58:82–115, 2020.

6.未来方向

  从数据、方法、应用及其相互作用的角度规划未来研究的可行方向

6.1新的数据资源探索

6.2基准数据库构建

6.3基于图形的工程知识应用

Robyn Speer, Joshua Chin, and Catherine Havasi. Conceptnet 5.5: An open multilingual graph of general knowledge. In Proceedings of the 31st AAAI Conference on Artificial IntelligenceFebruary, pages 4444–4451,San Francisco, California, USA, Feb. 4-9, 2017.
Bradley Camburn, Yuejun He, Sujithra Raviselvam, Jianxi Luo, and Kristin Wood. Machine learning-based design concept evaluation. Journal of Mechanical Design, 142(3), 2020.
Amrapali Zaveri, Anisa Rula, Andrea Maurino, Ricardo Pietrobon, Jens Lehmann, and Soeren Auer. Quality assessment for linked data: A survey. Semantic Web, 7(1):63–93, 2016.
Heiko Paulheim. Knowledge graph refinement: A survey of approaches and evaluation methods. Semantic web, 8(3):489–508, 2017.

6.4深度学习方法，新的模型，架构和XAI

  除5.3节写道的已经基于各种模型（如 CNn 和 RNN）的深度学习技术在支持 DbA 任务方面显示出了优势,应探索和利用最新的基础深度学习模型如：

transformer：Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser,and Illia Polosukhin. Attention is all you need. In Proceedings of the 31st Conference on Neural Information Processing Systems (NIPS), pages 5998–6008, Long Beach, CA, USA, Dec. 4-9, 2017.
GNN：Zonghan Wu, Shirui Pan, Fengwen Chen, Guodong Long, Chengqi Zhang, and S Yu Philip. A comprehensive survey on graph neural networks. IEEE transactions on neural networks and learning systems, 32(1):4–24, 2020.
BERT：Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding. Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), pages 4171–4186,2019.
GPT：Tom B Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, and Others. Language models are few-shot learners.The proceedings of 33th Conference on Neural Information Processing Systems (NeurIPS), pages 1877–1901,2020.

6.5数据驱动 DbA 的新子处理：基于类比的设计合成

  在传统的 DbA 过程中，按照类比检索和映射步骤，最终设计合成阶段通常需要设计师的经验、智慧和努力。与传统的设计合成策略（如形状语法和约束编程）不同，数据驱动的设计合成方法不一定需要专业知识，并且可以自动学会从数据集中生成合理的新设计。

机翼：
Wei Chen and Faez Ahmed. PaDGAN: Learning to Generate High-Quality Novel Designs. ASME Journal of Mechanical Design, 143(3):031703, 2021.
Wei Chen and Mark Fuge. Synthesizing Designs With Interpart Dependencies Using Hierarchical Generative Adversarial Networks. ASME Journal of Mechanical Design, 141(11):111403, 2019.
车轮：
Sangeun Oh, Yongsu Jung, Seongsin Kim, Ikjin Lee, and Namwoo Kang. Deep Generative Design: Integration of Topology Optimization and Generative Models. ASME Journal of Mechanical Design, 141(11):111405, 2019.
自行车：
Amin Heyrani Nobari, Muhammad Fathy Rashad, and Faez Ahmed. Creativegan: editing generative adversarial networks for creative design synthesis. In Proceedings of the 2021 ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE), Virtual, Online,Aug. 17-20, 2021.
设计材料微观结构：
Zijiang Yang, Xiaolin Li, L Catherine Brinson, Alok N Choudhary, Wei Chen, and Ankit Agrawal. Microstructural materials design via deep adversarial learning methodology. ASME Journal of Mechanical Design, 140(11):111416, 2018.
飞机：
Dule Shu, James Cunningham, Gary Stump, Simon W Miller, Michael A Yukish, Timothy W Simpson, and Conrad S Tucker. 3D Design Using Generative Adversarial Networks and Physics-Based Validation. ASME Journal of Mechanical Design, 142(7):071701, 2020.
社交机器人：
Yan Gan, Yingrui Ji, Shuo Jiang, Xinxiong Liu, Zhipeng Feng, Yao Li, and Yuan Liu. Integrating aesthetic and emotional preferences in social robot design: An affective design approach with Kansei Engineering and Deep Convolutional Generative Adversarial Network. International Journal of Industrial Ergonomics, 83:103128,2021.

  下图为与传统系统相比，对数据驱动的 DbA 系统进行概念化：

7.结论

  本文虽然只关注数据驱动的 DbA 方法和工具，但人工智能和数据科学技术对于增强其他经典工程设计方法也很有用。

Filippo Chiarello, Paola Belingheri, and Gualtiero Fantoni. Data science for engineering design : State of the art and future directions. Computers in Industry, 129:103447, 2021.
启发式设计：
Xiaoneng Jin, Mark Evans, Hua Dong, and Anqi Yao. Design heuristics for artificial intelligence: inspirational design stimuli for supporting UX designers in generating AI-powered ideas. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems, pages 1–8, Yokohama, Japan, May 8-13, 2021.
设计原理：
Katherine K Fu, Maria C Yang, and Kristin L Wood. Design Principles: Literature Review, Analysis, and Future Directions. ASME Journal of Mechanical Design, 138(10):101103, 2016.
产品族和平台设计：
Binyang Song, Jianxi Luo, and Kristin Wood. Data-driven platform design: Patent data and function network analysis. ASME Journal of Mechanical Design, 141(2):021101, 2019.
C-K：
Armand Hatchuel, Pascal Le Masson, Yoram Reich, and Eswaran Subrahmanian. Design theory: a foundation of a new paradigm for design science and engineering. Research in Engineering Design, 29(1):5–21, 2018.
组合设计：
Yuejun He. Combinational Creativity : Theories , Methods and Tools. PhD thesis, Singapore University of Technology and Design, 2019.