Method for parsing minimum equipment list manual and related device

By constructing a MEL hierarchy tree and context map to guide the large language model, the problems of professionalism and long text in the parsing of MEL manuals in the civil aviation field are solved, and efficient and accurate extraction of structured text data is achieved.

CN121328534BActive Publication Date: 2026-06-19CHINA EASTERN AIRLINES CO LTD +2
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHINA EASTERN AIRLINES CO LTD
Filing Date
2025-10-14
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing technologies struggle to efficiently and accurately parse MEL manuals in the civil aviation field, especially due to professional barriers, long text content, and insufficient reasoning capabilities of large language models, resulting in inconsistent outputs and low credibility.

Method used

By constructing a MEL hierarchy tree and context map, guidance is provided for large language models, reducing their comprehension difficulty and minimizing illusions. The MEL hierarchy tree provides a hierarchical relationship of item numbers, and the context map supplements attention deficits, transforming the task into a fill-in-the-blank exercise to improve output accuracy.

Benefits of technology

This improves the ability of large language models to understand MEL manuals and the reliability of their output results, reduces misidentification and illusions, and enhances parsing efficiency and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121328534B_ABST
    Figure CN121328534B_ABST
Patent Text Reader

Abstract

This disclosure relates to a method and related apparatus for parsing a Minimum Equipment List (MEL) manual. A method for parsing a Minimum Equipment List (MEL) manual includes: acquiring the MEL manual and prompt words; extracting raw text from at least a portion of the MEL manual; constructing a MEL hierarchy tree based on the MEL manual, the MEL hierarchy tree including multiple nodes with hierarchical relationships, each node corresponding to one of multiple item numbers in the MEL manual, the content of each node being initialized to empty; determining the context of each item number in the MEL manual based on the raw text to obtain a context map; inputting the MEL hierarchy tree, context map, and prompt words into a large language model, the large language model being configured to identify content in the context map corresponding to each node in the MEL hierarchy tree based on the prompt words and assign the corresponding content to that node; and outputting the processed MEL hierarchy tree via the large language model.
Need to check novelty before this filing date? Find Prior Art