A document conversion method, apparatus, device and medium

By combining an OCR engine and the BioBERT model with a medical terminology database for context-aware error correction, the problem of character errors and semantic integrity when converting PDF documents to Markdown using OCR is solved, achieving efficient and accurate document conversion.

CN122366356APending Publication Date: 2026-07-10GUANGZHOU FANGZHOU INFORMATION TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610456315.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-08
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

In existing technologies, when OCR recognizes PDF documents and converts them to Markdown format, there are problems such as character recognition errors, formatting errors, and difficulty in preserving semantic integrity. This is especially true in medical literature, where the cleaning algorithm is unstable.

Method used

An OCR engine is used to scan the document, recognize characters and calculate confidence scores. Context-aware error correction is performed using a medical terminology database and the BioBERT model. Error words are replaced using edit distance and semantic fit scores, and a document structure tree is constructed to generate a Markdown document.

Benefits of technology

It effectively reduces the rate of accidental deletion, preserves the semantic integrity of documents, improves processing efficiency and accuracy, and adapts to documents with different layout styles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122366356A_ABST
    Figure CN122366356A_ABST
Patent Text Reader

Abstract

This invention discloses a document conversion method, apparatus, device, and medium. This invention fully considers the specific context of words, and this context-aware error correction mechanism effectively avoids mechanical replacement without context, significantly reducing the false deletion rate and preserving the semantic integrity of the original document. Furthermore, this invention first recalls candidate words from a medical terminology database whose edit distance to the word to be verified is less than a distance threshold, and then calculates the semantic fit score. This two-stage strategy of "coarse screening + fine ranking" utilizes edit distance to quickly identify candidate words with similar shapes, and leverages the domain terminology database to ensure the effectiveness of candidate words within the medical semantic space. Compared to directly calculating semantic scores for all dictionary words, this method significantly reduces unnecessary semantic model inferences and improves processing efficiency; simultaneously, the domain specialization of the terminology database ensures that the recalled candidate words are highly relevant to the medical context.
Need to check novelty before this filing date? Find Prior Art