LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model

Traditional methods for requirement identification depend on the manual transformation of unstructured requirement texts into formal documents, a process that is both inefficient and prone to errors. Although requirement knowledge graphs offer structured representations, current named entity recogni...

Full description

Saved in:
Bibliographic Details
Main Authors: Yuhao Liu, Junjie Hou, Yuxuan Chen, Jie Jin, Wenyue Wang
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Aerospace
Subjects:
Online Access:https://www.mdpi.com/2226-4310/12/6/463
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849423871235588096
author Yuhao Liu
Junjie Hou
Yuxuan Chen
Jie Jin
Wenyue Wang
author_facet Yuhao Liu
Junjie Hou
Yuxuan Chen
Jie Jin
Wenyue Wang
author_sort Yuhao Liu
collection DOAJ
description Traditional methods for requirement identification depend on the manual transformation of unstructured requirement texts into formal documents, a process that is both inefficient and prone to errors. Although requirement knowledge graphs offer structured representations, current named entity recognition and relation extraction techniques continue to face significant challenges in processing the specialized terminology and intricate sentence structures characteristic of the aerospace domain. To overcome these limitations, this study introduces a novel approach for constructing aerospace-specific requirement knowledge graphs using a large language model. The method first employs the GPT model for data augmentation, followed by BERTScore filtering to ensure data quality and consistency. An efficient continual learning based on token index encoding is then implemented, guiding the model to focus on key information and enhancing domain adaptability through fine-tuning of the Qwen2.5 (7B) model. Furthermore, a chain-of-thought reasoning framework is established for improved entity and relation recognition, coupled with a dynamic few-shot learning strategy that selects examples adaptively based on input characteristics. Experimental results validate the effectiveness of the proposed method, achieving F1 scores of 88.75% in NER and 89.48% in relation extraction tasks.
format Article
id doaj-art-36a641ade98f45579c381295acc1bb6b
institution Kabale University
issn 2226-4310
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Aerospace
spelling doaj-art-36a641ade98f45579c381295acc1bb6b2025-08-20T03:30:25ZengMDPI AGAerospace2226-43102025-05-0112646310.3390/aerospace12060463LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language ModelYuhao Liu0Junjie Hou1Yuxuan Chen2Jie Jin3Wenyue Wang4China Aerospace Academy of Systems Science and Engineering, Beijing 100048, ChinaChina Aerospace Academy of Systems Science and Engineering, Beijing 100048, ChinaChina Aerospace Academy of Systems Science and Engineering, Beijing 100048, ChinaChina Aerospace Academy of Systems Science and Engineering, Beijing 100048, ChinaChina Aerospace Academy of Systems Science and Engineering, Beijing 100048, ChinaTraditional methods for requirement identification depend on the manual transformation of unstructured requirement texts into formal documents, a process that is both inefficient and prone to errors. Although requirement knowledge graphs offer structured representations, current named entity recognition and relation extraction techniques continue to face significant challenges in processing the specialized terminology and intricate sentence structures characteristic of the aerospace domain. To overcome these limitations, this study introduces a novel approach for constructing aerospace-specific requirement knowledge graphs using a large language model. The method first employs the GPT model for data augmentation, followed by BERTScore filtering to ensure data quality and consistency. An efficient continual learning based on token index encoding is then implemented, guiding the model to focus on key information and enhancing domain adaptability through fine-tuning of the Qwen2.5 (7B) model. Furthermore, a chain-of-thought reasoning framework is established for improved entity and relation recognition, coupled with a dynamic few-shot learning strategy that selects examples adaptively based on input characteristics. Experimental results validate the effectiveness of the proposed method, achieving F1 scores of 88.75% in NER and 89.48% in relation extraction tasks.https://www.mdpi.com/2226-4310/12/6/463system engineeringrequirement identificationlarge language modelknowledge graphnamed entity recognitionrelation extraction
spellingShingle Yuhao Liu
Junjie Hou
Yuxuan Chen
Jie Jin
Wenyue Wang
LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
Aerospace
system engineering
requirement identification
large language model
knowledge graph
named entity recognition
relation extraction
title LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
title_full LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
title_fullStr LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
title_full_unstemmed LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
title_short LLM-ACNC: Aerospace Requirement Texts Knowledge Graph Construction Utilizing Large Language Model
title_sort llm acnc aerospace requirement texts knowledge graph construction utilizing large language model
topic system engineering
requirement identification
large language model
knowledge graph
named entity recognition
relation extraction
url https://www.mdpi.com/2226-4310/12/6/463
work_keys_str_mv AT yuhaoliu llmacncaerospacerequirementtextsknowledgegraphconstructionutilizinglargelanguagemodel
AT junjiehou llmacncaerospacerequirementtextsknowledgegraphconstructionutilizinglargelanguagemodel
AT yuxuanchen llmacncaerospacerequirementtextsknowledgegraphconstructionutilizinglargelanguagemodel
AT jiejin llmacncaerospacerequirementtextsknowledgegraphconstructionutilizinglargelanguagemodel
AT wenyuewang llmacncaerospacerequirementtextsknowledgegraphconstructionutilizinglargelanguagemodel