A survey of LLM-augmented knowledge graph construction and application in complex product design
Xinxin Liang, Zuoxu Wang, Mingrui Li, Zhijie Yan
Abstract
In the field of complex product design, deploying knowledge graphs (KGs) has become a promising trend due to its strength on exploiting and applying the large-scale, complex, and specialized domain knowledge. In recent years, large language models (LLMs) have also attracted much attention due to their outstanding performance in natural language understanding and generation. However, in the research of complex product design dominated by domain knowledge, few studies involve LLMS and KG at the same time. To fill this gap, we survey 42 articles published in the last four years, focusing on three key questions. The combination of LLM and KG in specific applications in complex product design is deeply discussed. The analysis reveals how these techniques facilitate data collection, design concept formation and design process optimization and proposes a technical framework combining LLM and KG for complex product design domain. In addition, we identify key challenges and propose directions for future research. As an explorative survey paper, this paper provides insightful ideas for implementing more specialized domain knowledge graph in complex product design field.