Knowledge Base Repository
In addition to research papers, the Design Society is developing several valuable resources for those interested in the study of design. These include a repository of PhD theses, a library of case studies and transcripts of design activities, and an archive of our newsletters. Please note that these resources are accessible exclusively to Design Society members.
From Text to Structure: Extracting and Validating Complex System Representations Using Large Language Models
Carlo Lipizzi
Type:
Year:
2025
Editor:
Ali A. Yassine, Christopher Langner, Matthias Kreimeyer, Tyson R. Browning, Steven D. Eppinger
Author:
Series:
DSM
Institution:
Stevens Institute of Technology, United States of America
Page(s):
145-153
DOI number:
Abstract:
Managing complexity in systems engineering requires accurate representations of the system's structure and dependencies. This work introduces a semi-automated/human-in-the-loop method to extract and validate structural representations of complex systems from textual descriptions using Large Language Models (LLMs). The proposed pipeline consists of: (1) collecting relevant system documentation, (2) extracting subject–predicate–object triplets using an LLM, (3) building a graph where nodes are the subjects and the objects - representing system entities - and edges are the predicates - representing relationships, and (4) conducting Subject Matter Expert (SME) validation of the resulting graph. This verified graph serves as a formal structural model of the system. The graph could be used as such as well as to assess the consistency of future LLM-generated responses.
Keywords:
