Return
LLM-based domain-specific knowledge graph construction and hybrid retrieval-augmented generation
X
S
Y
W
C
L
DOI:10.1016/j.jmsy.2026.04.023.png)
Abstract
En 中文
• An intent-constrained hybrid RAG framework is proposed for knowledge-intensive industrial scenarios. • A domain-specific knowledge extraction method integrating text classification and LLMs is developed. • An intent-constrained hybrid retrieval algorithm with subgraph expansion enables structured evidence organization. • The framework is validated on superplastic forming equipment, improving retrieval quality and decision support.
Keywords:
domain-specific knowledge graph
hybrid retrieval-augmented generation
intent-constrained framework
industrial scenarios
large language models
Journal
IF:
14.2
Papers:
2.6K
Citations:
1.6W
