基于混合模型的外事差旅证件结构化信息提取方法及系统
By using a hybrid model approach, attribute names and word segmentation algorithms are used to split the text of foreign affairs travel documents. By combining lexical cosine similarity and structural priority algorithms, the problem of inaccurate information extraction caused by differences in document format is solved, and efficient structured processing of document information is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING DEXUN AVIATION SERVICE CO LTD
- Filing Date
- 2025-12-10
- Publication Date
- 2026-07-17
AI Technical Summary
The information extraction of different foreign affairs travel documents varies greatly in format, resulting in low accuracy and reliability of information extraction. Traditional methods are difficult to adapt to the different formats of various documents, leading to information overlap and reduced recognition accuracy.
A hybrid model-based approach is adopted, which splits the document text sequence by attribute names and word segmentation algorithms, and combines lexical cosine similarity and structural priority algorithms to determine core words and effective word combinations, thereby constructing a well-structured information with clear hierarchy.
It improves the accuracy and reliability of retrieving foreign affairs travel document information, meeting the needs of foreign affairs operations for rapid and accurate access to document information.
Smart Images
Figure CN122021615B_ABST