设计决策脉络挖掘方法  

Design Decision Context Mining

作　　者：刘继红[1] 王克剑[1,2] 王嘉骥 Liu Jihong;Wang Kejian;Wang Jiaji(School of Mechanical Engineering & Automation,Beihang University,Beijing 100191;Harbin Electric Company Limited,Harbin 150028)

机构地区：[1]北京航空航天大学机械工程及自动化学院,北京100191 [2]哈尔滨电气股份有限公司,哈尔滨150028

出　　处：《计算机辅助设计与图形学学报》2019年第6期1053-1060,共8页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(51475027,51575046)

摘　　要：为了更好地引导和启发设计者的思考,提出设计决策脉络概念.设计决策脉络是舍去设计试错过程并呈现设计主体思维因果关系发展路径的设计理性知识模型抽象表达.基于细粒度设计理性知识模型,提出设计决策脉络挖掘方法.该方法基于商空间理论对设计理性知识模型进行约简,构建分层递阶设计决策脉络模型,以支持设计者从不同粒度水平分析设计问题;通过改进的流形排序算法实现模型的语义约简,删除与设计主题相关度较低的知识片段以获取核心设计思考过程.最后通过构建原型系统,验证了文中方法是有效的.In order to guide and inspire the designer’s thinking efficiently, this paper puts forward the concept of design decision context, which is an abstract expression of design rationale knowledge models by ignoring design trial-and-error processes and iterations. Design decision context is the core causality development path of design thinking. Based on the proposed fine-grained design rationale knowledge model, this paper proposes a design decision context mining method. Firstly, a hierarchical design decision model is constructed based on the quotient space theory, which can support the designer to analyze the design problem from different levels of granularity. Secondly, semantic reduction of the model is realized through the improved manifolds ranking method. Finally, the core design thinking process, i.e., design decision context, is obtained by eliminating the less relevant knowledge fragments. The proposed mining method is verified by a developing a prototype system.

关 键 词：设计理性 设计决策脉络 商空间理论 流形排序 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]