《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (8): 2681-2690.DOI: 10.11772/j.issn.1001-9081.2025091103

• 前沿与综合应用 • 上一篇    

面向航空领域的消息一致性解析框架

王梦婷1, 高拴梁2(), 廖文睿2, 李赓2, 张嘉兴2, 祝远芳2, 赵天歌1, 刘勇慧1, 梁红茹2   

  1. 1.中国人民解放军93216部队,北京 100085
    2.四川大学 计算机学院,成都 610065
  • 收稿日期:2025-09-22 修回日期:2025-11-06 接受日期:2025-11-07 发布日期:2025-12-04 出版日期:2026-08-10
  • 通讯作者: 高拴梁
  • 作者简介:王梦婷(1993—),女,北京人,工程师,主要研究方向:卫星通信
    高拴梁(2000—),男,内蒙古乌兰察布人,硕士研究生,主要研究方向:自然语言处理
    廖文睿(2002—),男,四川威远人,硕士研究生,主要研究方向:流程自动化
    李赓(2003—),男,山东聊城人,主要研究方向:自然语言处理
    张嘉兴(2004—),男,河北唐山人,主要研究方向:多模态大语言模型
    祝远芳(2004—),女,江西新余人,主要研究方向:自然语言处理
    赵天歌(1992—),女,辽宁沈阳人,工程师,硕士,主要研究方向:航空通信
    刘勇慧(1990—),男,内蒙古赤峰人,助理工程师,硕士,主要研究方向:航空通信
    梁红茹(1992—),女,河北新乐人,副教授,博士,主要研究方向:自然语言处理。
  • 基金资助:
    国家自然科学基金面上项目(62576230);多域数据协同处理与控制全国重点实验室开放基金资助项目(CLDL-20240201)

Message consistency parsing framework for aviation domain

Mengting WANG1, Shuanliang GAO2(), Wenrui LIAO2, Geng LI2, Jiaxing ZHANG2, Yuanfang ZHU2, Tiange ZHAO1, Yonghui LIU1, Hongru LIANG2   

  1. 1.Unit 93216 of the Chinese People’s Liberation Army,Beijing 100085,China
    2.College of Computer Science,Sichuan University,Chengdu Sichuan 610065,China
  • Received:2025-09-22 Revised:2025-11-06 Accepted:2025-11-07 Online:2025-12-04 Published:2026-08-10
  • Contact: Shuanliang GAO
  • About author:WANG Mengting, born in 1993, engineer. Her research interests include satellite communication.
    LIAO Wenrui, born in 2002, M. S. candidate. His research interests include process automation.
    LI Geng, born in 2003. His research interests include natural language processing.
    ZHANG Jiaxing, born in 2004. His research interests include multimodal large language models.
    ZHU Yuanfang, born in 2004. Her research interests include natural language processing.
    ZHAO Tiange, born in 1992, M. S., engineer. Her research interests include aviation communication.
    LIU Yonghui, born in 1990, M. S., assistant engineer. His research interests include aviation communication.
    LIANG Hongru, born in 1992, Ph. D., associate professor. Her research interests include natural language processing
  • Supported by:
    General Program of National Natural Science Foundation of China(62576230);Open Fund of State Key Laboratory of Multi-domain Data Collaborative Processing and Control(CLDL-20240201)

摘要:

在航空领域,为了确保空中单位与地面单位之间通信的稳定性,非结构化的文本消息需被转化为固定格式的结构化信息数据进行传输。现有方法虽利用大语言模型(LLM)的指令遵循能力实现从文本到结构化数据的转换,却普遍忽视了LLM固有的幻觉问题与不可解释性对航空信息安全构成的严重威胁。因此,提出面向航空领域的消息一致性解析框架(MCPFA),通过融合模型自我反思与人机协同机制,系统性地保障信息转化过程中的语义保真与操作安全。首先,提出基于自我反思的一致性检验方法(CVMSR),从语法、语义和语用这3个维度构建递进式验证机制,实现对结构化输出的动态校验与修正;其次,设计智能化航空消息人机协作检验系统(IAMHCIS),通过置信度感知与跨层一致性分析精准触发专家干预,并将人类反馈融入模型迭代,形成闭环优化;最后,通过系统性实验全面评估框架性能。实验结果表明,MCPFA在一致性判断与修正任务上显著优于现有方法,不仅能有效识别并修复因术语混淆、单位误判或规章违背导致的关键错误,还在未见过的消息类型上展现出良好的泛化能力。更重要的是,引入专家知识后,模型性能持续提升,从而验证了所提框架的可成长性与工程实用价值,为航空等安全攸关领域的LLM应用提供了兼顾准确性、安全性与可解释性的新路径。

关键词: 信息抽取, 大语言模型, 模型自我反思, 人机协作, 幻觉

Abstract:

In aviation domain, stable communication between air units and ground units requires unstructured text messages be transformed into structured information in a fixed format for transmission. Although the existing methods use the instruction-following capabilities of Large Language Models (LLMs) to perform such transformation from text into structured data, they ignore the inherent hallucination and lack of interpretability of LLMs — which pose serious threats to information security. Therefore, a Message Consistency Parsing Framework for the Aviation domain (MCPFA) was proposed to integrate model self-reflection and human-machine collaboration to ensure semantic fidelity and operational safety during the parsing process systematically. First, a Consistency Verification Method based on Self-Reflection (CVMSR) was developed to establish a progressive validation mechanism across three dimensions: syntax, semantics, and pragmatics, so as to realize dynamic verification and correction of the structured output. Second, an Intelligent Aviation Message Human-Machine Collaboration Inspection System (IAMHCIS) was designed to use confidence awareness and cross-layer consistency analysis to trigger expert intervention precisely and incorporate human feedback into model iteration, so as to form a closed-loop optimization. Finally, the framework was evaluated comprehensively through systematic experiments. The results demonstrate that MCPFA outperforms the existing methods significantly in both consistency judgment and correction tasks, identifying and fixing critical errors caused by term confusion, unit misinterpretation, or regulation violation effectively. Furthermore, the framework exhibits strong generalization capability on unseen message types. Importantly, after integrating expert knowledge, performance gains continuously, confirming the proposed framework’s growability and practical value, and the proposed framework provides a new accurate, secure, and interpretable way for LLM application in aviation and other safety-critical domains.

Key words: information extraction, Large Language Model (LLM), model self-reflection, human-machine collaboration, hallucination

中图分类号: