Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge

Electrical standards encompass complex technical requirements across multiple disciplines, making their management and application a significant challenge that urgently requires efficient solutions. This paper proposes a knowledge graph retrieval-enhanced training method for large language models (...

Full description

Saved in:
Bibliographic Details
Main Authors: Sai Zhang, Xiaoxuan Fan, Bochuan Song, Xiao Liang, Qiang Zhang, Zhihao Wang, Bo Zhang
Format: Article
Language:English
Published: Ital Publication 2025-06-01
Series:HighTech and Innovation Journal
Subjects:
Online Access:https://hightechjournal.org/index.php/HIJ/article/view/969
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849244577684258816
author Sai Zhang
Xiaoxuan Fan
Bochuan Song
Xiao Liang
Qiang Zhang
Zhihao Wang
Bo Zhang
author_facet Sai Zhang
Xiaoxuan Fan
Bochuan Song
Xiao Liang
Qiang Zhang
Zhihao Wang
Bo Zhang
author_sort Sai Zhang
collection DOAJ
description Electrical standards encompass complex technical requirements across multiple disciplines, making their management and application a significant challenge that urgently requires efficient solutions. This paper proposes a knowledge graph retrieval-enhanced training method for large language models (LLMs). By leveraging a pre-trained language model (PLM), highly similar subgraphs are retrieved from the electrical standards knowledge graph. These subgraphs are then parsed into triples using entity linking and semantic reasoning. The triples are converted into natural language text by the LLM, which combines them with the input question to perform reasoning and generate accurate answers. The proposed method addresses the complexity of question answering for electrical standards and offers a novel approach for managing and applying these standards in the field of electrical engineering. Experimental results demonstrate that this approach significantly enhances the model's understanding of electrical standards, enabling it to generate more accurate answers.
format Article
id doaj-art-010e05eb5b73494bac9aa45e96b7918e
institution Kabale University
issn 2723-9535
language English
publishDate 2025-06-01
publisher Ital Publication
record_format Article
series HighTech and Innovation Journal
spelling doaj-art-010e05eb5b73494bac9aa45e96b7918e2025-08-20T03:59:08ZengItal PublicationHighTech and Innovation Journal2723-95352025-06-016210.28991/HIJ-2025-06-02-05Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard KnowledgeSai Zhang0https://orcid.org/0009-0000-0963-8893Xiaoxuan Fan1https://orcid.org/0009-0006-2463-8983Bochuan Song2Xiao Liang3Qiang Zhang4Zhihao Wang5Bo Zhang6State Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Laboratory of Grid Advanced Computing and Applications, China Electric Power Research Institute Co., Ltd., BeijingState Grid Wuxi Power Supply Company of Jiangsu Electric Power Co., Ltd., Wuxi Electrical standards encompass complex technical requirements across multiple disciplines, making their management and application a significant challenge that urgently requires efficient solutions. This paper proposes a knowledge graph retrieval-enhanced training method for large language models (LLMs). By leveraging a pre-trained language model (PLM), highly similar subgraphs are retrieved from the electrical standards knowledge graph. These subgraphs are then parsed into triples using entity linking and semantic reasoning. The triples are converted into natural language text by the LLM, which combines them with the input question to perform reasoning and generate accurate answers. The proposed method addresses the complexity of question answering for electrical standards and offers a novel approach for managing and applying these standards in the field of electrical engineering. Experimental results demonstrate that this approach significantly enhances the model's understanding of electrical standards, enabling it to generate more accurate answers. https://hightechjournal.org/index.php/HIJ/article/view/969Electric Standards KnowledgeLLMRAGKnowledge GraphSemantic Reasoning
spellingShingle Sai Zhang
Xiaoxuan Fan
Bochuan Song
Xiao Liang
Qiang Zhang
Zhihao Wang
Bo Zhang
Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
HighTech and Innovation Journal
Electric Standards Knowledge
LLM
RAG
Knowledge Graph
Semantic Reasoning
title Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
title_full Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
title_fullStr Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
title_full_unstemmed Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
title_short Research on RAG-Based Cognitive Large Language Model Training Method for Power Standard Knowledge
title_sort research on rag based cognitive large language model training method for power standard knowledge
topic Electric Standards Knowledge
LLM
RAG
Knowledge Graph
Semantic Reasoning
url https://hightechjournal.org/index.php/HIJ/article/view/969
work_keys_str_mv AT saizhang researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT xiaoxuanfan researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT bochuansong researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT xiaoliang researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT qiangzhang researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT zhihaowang researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge
AT bozhang researchonragbasedcognitivelargelanguagemodeltrainingmethodforpowerstandardknowledge