Research on design support by creative abduction

创意溯因的设计支持研究

基本信息

批准号：
14380170
负责人：
TAKEDA Hideaki
金额：
$ 7.1万
依托单位：
National institute of Informatics
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (B)
财政年份：
2002
资助国家：
日本
起止时间：
2002 至 2004
项目状态：
已结题

项目摘要

This project presents "Universal Abduction Studio" (UAS) that is a new approach to computational support for designers and discusses technologies to implement it. Although design support systems have been well developed for geometric and detail design stages, they are still unsuccessful in the conceptual design stage. In our opinion, the main difficulty comes from incomplete and insufficient understanding about design knowledge and its operations that play a crucial role in conceptual design. In contrast, however, recently thanks to development of the Internet technologies, more and more knowledge is accumulated and available electronically. It then becomes an interesting research question how to apply such an enormous amount of diverse knowledge to conceptual design. UAS is not intended to be an automated design system. UAS is an environment in which designers can manage and apply knowledge to arrive at new, creative design collaboratively with the system. UAS provides the designer wi … More th knowledge and reasoning functions, while the designer selects operations and evaluates design results iteratively. we show a novel knowledge representation to support creative design process. Our RDF-based representation consists of modules for knowledge repositories like predicate and object ontologies. A user can compose an individual model for knowledge by selecting and combining suitable elements in these knowledge repositories. This representation solves two characteristics observed in our case study analysis, ie., multiplicity of interpretation and variety of knowledge. The graph structure can generate multiple interpretations with minimum modification such as redirecting "if" and "then" links. Furthermore, since we separate metamodel and aspect-specific models, we can obtain various aspect-specific models from a single metamodel. The problem to handle with variety of knowledge is balance of expansivity and consistency. Since we provide categories of knowledge like object and predicate ontologies in order to clarify roles of knowledge, we can add easily new knowledge like domain-specific knowledge according to its roles without avoiding messy. Less

该项目提出了“Universal Abduction Studio”（UAS），这是一种为设计师提供计算支持的新方法，并讨论了实现它的技术，尽管设计支持系统在几何和详细设计阶段已经得到了很好的发展，但在概念上仍然不成功。我们认为，主要的困难来自于对在概念设计中起着至关重要作用的设计知识及其操作的不完整和不充分的理解，然而，最近由于互联网技术的发展，越来越多的知识被忽视。然后它就成为一个有趣的研究问题。如何将如此大量的多样化知识应用于概念设计 UAS 并不是一个自动化设计系统，设计人员可以在其中管理和应用知识以与系统协作实现新的创意设计。为设计者提供知识和推理功能，而设计者可以迭代地选择操作并评估设计结果，我们展示了一种新颖的知识表示来支持创意设计过程。我们基于 RDF 的表示由谓词和对象等知识存储库模块组成。用户可以通过选择和组合这些知识库中的合适元素来构建单独的知识模型，这种表示解决了我们在案例研究分析中观察到的两个特征，即解释的多样性和知识的多样性。具有最小修改的多种解释，例如重定向“if”和“then”链接此外，由于我们将元模型和特定于方面的模型分开，因此我们可以从单个元模型中获得各种特定于方面的模型来处理各种知识的问题。是平衡由于我们提供了诸如对象和谓词本体之类的知识类别来阐明知识的角色，因此我们可以根据其角色轻松添加新知识（例如特定于领域的知识），而不会避免混乱。