Creativity is an important skill for design teams to reach new and useful solutions. Designers often use one or more of creativity and innovation techniques (CITs) to achieve the desired creative potential during new product development (NPD). The selection of adequate CITs requires considerable expertise, given the multiple application contexts and the extensive number of techniques available. The purpose of this study is to present a creativity support system able to manage this amount of information and provide valuable knowledge to improve NPD.,This study presents a knowledge-based system prototype using artificial intelligence (AI) to support knowledge management on the selection of CITs for design. CITs assertion is modelled through a double inference process using five categories, correlating over 500 different entry scenarios to 24 implemented CITs. The techniques are classified according to: design stage, innovation focus, team relationship, execution method and difficult of use. Prototype outputs explanations on the inference process and chosen techniques information.,To demonstrate the system scope, two opposite design cases are presented. The system was validated by experts in knowledge management and mechanical engineering design. The validation process demonstrates relevance of the approach and improvement directions for future developments.,Though literature contains toolkits and taxonomy for CITs, no work applies AI to identify design scenarios, select best CITs and instruct about their use. Validators reported to know less than half of the available techniques, showing a clear knowledge gap among design experts.

中文翻译：

---

一种支持知识管理的人工智能方法，用于选择创造力和创新技术

创造力是设计团队获得新的有用解决方案的一项重要技能。设计师经常使用一种或多种创造力和创新技术（CIT）在新产品开发（NPD）期间实现所需的创造潜力。考虑到多种应用环境和大量可用技术，选择合适的CIT需要大量的专业知识。这项研究的目的是提供一个能够支持这种信息量并提供有价值的知识以改善NPD的创造力支持系统。该研究提出了一种使用人工智能（AI）来支持选择知识管理的基于知识的系统原型。设计的CIT。CIT断言是通过使用五个类别的双重推理过程建模的，将500多种不同的进入场景与24个已实施的CIT相关联。这些技术根据以下几个方面进行分类：设计阶段，创新重点，团队关系，执行方法和使用难度。原型输出有关推理过程的说明和所选技术信息。为了演示系统范围，提出了两个相反的设计案例。该系统已通过知识管理和机械工程设计方面的专家验证。验证过程证明了该方法和未来开发的改进方向的相关性。尽管文献中包含CIT的工具包和分类法，但没有工作将AI用于识别设计方案，选择最佳CIT并指导其使用。验证者报告了解的可用技术不到一半，这表明设计专家之间存在明显的知识鸿沟。