一种文档体裁的识别方法、装置、设备及存储介质
By acquiring summary information and chapter content of documents, and using a pre-trained model to identify document genres, the accuracy and long-tail problems of document genre identification are solved, achieving more accurate document classification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING ZITIAO NETWORK TECH CO LTD
- Filing Date
- 2023-09-19
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, document genre recognition is greatly affected by subjective factors, has low accuracy, and suffers from the problem of long-tail genre labeling, resulting in insufficient or excessive genre labeling.
By acquiring summary information and chapter content of the target document, and using a pre-trained genre recognition model, combined with the title, summary, and a set number of chapter contents, the document genre is identified. The mapping relationship between the title, summary, and chapter contents is considered to increase the number of genres that can be identified.
It improves the accuracy of document genre recognition, solves the long-tail problem of genres, increases the number of genres identified, and enhances the effect of document classification.
Smart Images
Figure CN117271769B_ABST