Data mining prototyping knowledge graphs for design process insights
Data mining prototyping knowledge graphs for design process insights
复制标题
DOI：
10.1080/09544828.2024.2302746
复制
发表时间：
2024-01
影响因子：
2.7
通讯作者：
J. Gopsill;L. Giunta;M. Goudswaard;C. Snider;B. Hicks
中科院分区：
文献类型：
--
作者：
J. Gopsill;L. Giunta;M. Goudswaard;C. Snider;B. Hicks
原型知识捕获的进步导致了几个基于图的模式，其中节点表示原型活动和个人，边表示设计团队定义的原型之间的语义关系（例如影响和参与）。他们随后在基于Web的应用程序中的实现，使物理和虚拟原型知识的集成。该集成提供了可追溯性，版本控制和图形分析的应用，以生成对工程设计过程的见解，并且先前研究中报告的模式验证研究已经产生了几个开放访问的原型知识图谱数据集。本文建立在以前的研究，通过贡献九个原型知识图的数据挖掘分析。分析遵循CRISP-DM过程。其目的是在设计过程中的原型行为的见解。结果显示了9项发现。5项证实了现有的意见，4项是新的意见。特别是，数据挖掘的知识图显示序列和类型的原型生成的知识是唯一的每个设计过程。此外，关键设计路径由物理原型主导。
Advancements in prototyping knowledge capture have resulted in several graph-based schemas where nodes represent prototyping activities and individuals, and edges represent semantic relationships between prototypes (e.g. influence and involvement) defined by the design team. Their subsequent implementation in web-based applications has enabled the integration of both physical and virtual prototyping knowledge. The integration affords traceability, version control, and application of graph analysis to generate insights into the Engineering Design process and the schema validation studies reported in prior research have resulted in several open-access Prototyping Knowledge Graph datasets. This paper builds on prior research by contributing a data mining analysis of nine Prototyping Knowledge Graphs. The analysis followed the CRISP-DM process. The purpose was to elicit insights on prototyping behaviour within the design process. The results revealed nine findings. Five corroborated existing observations and four were new observations. In particular, data mining of the knowledge graphs showed sequence and type of knowledge generated from prototyping are unique to each design process. Also, the critical design path is dominated by physical prototypes.