수명주기관리계획서의 선별적 개정을 위한 LLM-치환규칙 기반 문서 처리 방안 연구

박재성¹^{, *} ; 이승률² ; 정준² ; 오준혁³

1LIG Defense&Aerospace 미래전장IPS연구소 연구원
2LIG Defense&Aerospace 미래전장IPS연구소 수석연구원
3LIG Defense&Aerospace 미래전장IPS연구소 선임연구원

Study on an LLM and Replacement Rule-Based Document Processing Approach for Selective Revision of Life Cycle Sustainment Plan

Jae-Seong Park¹^{, *} ; Seung-Ryool Lee² ; Jun Jeong² ; Jun-Hyeok Oh³

1Research engineer, Future Warfare Integrated Product Support R&D (Maritime), LIG Defense&Aerospace
2Chief research engineer, Future Warfare Integrated Product Support R&D (Maritime), LIG Defense&Aerospace
3Research engineer, Future Warfare Integrated Product Support R&D (Maritime), LIG Defense&Aerospace

Correspondence to: *Jae-Seong Park 20, Daewangpangyo-ro 851beon-gil, Sujeong-gu, Seongnam-si, Gyeonggi-do, 13449, Republic of Korea Tel: +82-31-8038-0592 E-mail: jaeseong.park01@ligdna.com

초록

본 논문에서는 기존 사업의 수명주기관리계획서를 신규사업 정보에 맞게 개정하기 위한 LLM 및 치환규칙 기반 문서 처리 방안을 제안한다. 수명주기관리계획서는 사업별 정보가 반복적으로 포함되는 장문 기술문서로 기존 문서 구조를 유지하면서 필요한 항목을 일관되게 수정할 필요가 있다. 그러나 LLM 단독으로 문서를 직접 재작성하는 경우 원문과 관련 없는 내용이 생성되거나 문서 구조가 변형될 가능성이 있다. 이에 본 연구에서는 LLM을 개정 후보 탐색 및 제안 수단으로 활용하고, 사용자가 승인한 치환규칙만 문서에 반영하는 방식을 구성하였다. 기존 문서 구조를 유지하면서 개정 대상 표현을 선별적으로 반영하고, 변경 이력과 잔존 후보를 검토할 수 있는 문서 처리 절차를 제시하였다.

Abstract

This study proposes a document processing method based on an LLM and replacement rules for revising an existing life cycle sustainment plan using new project information. A life cycle sustainment plan is a long-form technical document that repeatedly contains project-specific information. Therefore, necessary items must be revised consistently while preserving the structure of the existing document. However, when an LLM directly rewrites the document, it may generate content unrelated to the source document or alter the original document structure. To address this issue, this study used an LLM to identify and suggest revision candidates and applied only user-approved replacement rules to the document. The proposed method selectively reflects expressions requiring revision while preserving the existing document structure and supports the review of change history and residual candidates.

Keywords:

Large Language Model, Life Cycle Sustainment Plan, Replacement Rule, Selective Revision, Document Processing

키워드:

대규모 언어모델, 수명주기관리계획서, 치환규칙, 사용자 승인, 문서 처리

References

Ministry of National Defense, Directive on Total Life Cycle Management, MND Directive No. 3144, National Law Information Center, 2026.
Defense Acquisition Program Administration, Regulation on Life Cycle Management in the Acquisition Phase, DAPA Directive No. 960, National Law Information Center, 2026.
Ziwei Ji, Nayeon Lee, Rita Frieske, Tiezheng Yu, Dan Su, Yan Xu, Etsuko Ishii, Ye Jin Bang, Andrea Madotto, & Pascale Fung, ‘Survey of Hallucination in Natural Language Generation,’ ACM Computing Surveys, VOL. 55, NO. 12, 2023, pp. 1-38. [https://doi.org/10.1145/3571730]
Joshua Maynez, Shashi Narayan, Bernd Bohnet, & Ryan McDonald, ‘On Faithfulness and Factuality in Abstractive Summarization,’ in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, July 2020, pp. 1906-1919. [https://doi.org/10.18653/v1/2020.acl-main.173]
Defense Acquisition University, Integrated Product Support (IPS) Elements Guidebook, Version 1.0b, 2024.
Defense Acquisition Program Administration, Standard Special Contract Conditions for Naval Shipbuilding (General and Defense Industry), DAPA Established Rule No. 1075, National Law Information Center, 2026.
Eduardo Mosqueira-Rey, Elena Hernández-Pereira, David Alonso-Ríos, José Bobes-Bascarán, & Ángel Fernández-Leal, ‘Human-in-the-Loop Machine Learning: A State of the Art,’ Artificial Intelligence Review, VOL. 56, 2023, pp. 3005-3054. [https://doi.org/10.1007/s10462-022-10246-w]
Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin et al., ‘A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions,’ ACM Transactions on Information Systems, VOL. 43, NO. 2, 2025, Article 42. [https://doi.org/10.1145/3703155]
Konstantinos Lazaros, Aristidis G. Vrahatis, & Sotiris Kotsiantis, ‘Human-in-the-Loop Artificial Intelligence: A Systematic Review of Concepts, Methods, and Applications,’ Entropy, VOL. 28, NO. 4, 2026, Article 377. [https://doi.org/10.3390/e28040377]
Ben Shneiderman, ‘Human-Centered Artificial Intelligence: Reliable, Safe & Trustworthy,’ International Journal of Human-Computer Interaction, VOL. 36, NO. 6, 2020, pp. 495-504. [https://doi.org/10.1080/10447318.2020.1741118]