• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2017, Vol. 53 ›› Issue (15): 32-39.doi: 10.3901/JME.2017.15.032

• 特邀专栏:设计思维及在创新设计中的应用 • 上一篇    下一篇

基于语义链接的设计认知多维建模方法

徐江1,2, 王修越3, 王奕3, 欧细凡1   

  1. 1. 同济大学设计创意学院 上海 200092;
    2. 同济大学上海国际设计创新研究院 上海 200092;
    3. 东南大学-蒙纳士大学苏州联合研究生院 苏州 235000
  • 出版日期:2017-08-05 发布日期:2017-08-05
  • 作者简介:徐江,男,1978年生,博士,副教授。主要研究方向为国家创新设计发展战略、设计认知计算、情感人机工程等。E-mail:xujzju@seu.edu.cn;王修越,男,1990年出生,硕士。主要研究方向为设计认知计算、设计信息可视化等。E-mail:wxyseu@qq.com
  • 基金资助:
    国家自然科学基金(51675382,51205059)、中国工程院重点咨询项目(2015-ZD-15)、上海市设计学IV类高峰学科资助项目(DA17001,DB17001)和上海市教育发展基金会晨光计划(15CG72)资助项目

The Multidimensional Modeling Method of Design Cognition Based on Semantic Link

XU Jiang1,2, WANG Xiuyue3, WANG Yi3, OU Xifan1   

  1. 1. College of Design and Innovation, Tongji University, Shanghai 200092;
    2. Shanghai International Collge of Design and Innovation, Tongji University, Shanghai 200092;
    3. Southeast University and Monash University Joint Graduate School, Suzhou 235000
  • Online:2017-08-05 Published:2017-08-05

摘要: 针对复杂设计认知过程及其概念演化研究的模糊与不确定问题,从设计认知语义推理及链接关系表征出发,以设计认知沟通文本为基础,建立集成数据链接、过程链接和概念链接的多维设计认知模型。该模型采用矩阵存储语义链接数据,处理非结构化设计认知数据,进而融合确定性信息理论T-code编码方法与链接表求取链接节点熵值,评判设计认知过程的复杂性,实现关键创意概念捕获和剖析设计意图转换。在非时序化的概念空间关系构建方面,基于语义网络的聚类分析设计认知的概念演化,并回溯提取演化路径。以此为基础,运用实例分析验证多维建模方法的有效性,开发设计认知可视化辅助分析工具,从多维度提供设计修改和评估依据,提升设计认知计算研究效率。

关键词: 多维建模, 链接表, 设计认知, 语义链接

Abstract: The complex design cognition and concept evolution limited to process research, is involved in the problem of ambiguity and uncertainty analysis. From the perspective of cognitive semantic reasoning and link representation, a deep mining on understanding design process is conducted by using the semantic links as the representation of design cognition information. On the basis of these, a multi-dimensional cognitive model which integrates data links, process links and conceptual links is proposed by emphasizing on the cognitive communication text information. The model using matrix to storage semantic link data processes unstructured data of design cognition. Though analyzing the link node entropy obtained by merging the uncertain information theory, T-code coding method and linkography, the complexity of the design cognition process is evaluated to realize the capture of key innovative concepts and analysis of switching design intention. On the aspect of non-sequential concept space relations, the semantic network clusters focus on analyzing the evolution of cognitive concept and extracting the evolutionary path. Furthermore, effectiveness on the multi-dimensional model is proved through case study on design experiments realizing a language data driven research method. Linkage software is also developed to realize the automatic computation and data visualization aiming at providing design modification and evaluation basis from multi dimensions and improving the automation and efficiency on design cognition and computation.

Key words: design cognition, linkography, multi-dimensional cognitive model, semantic link

中图分类号: