README.md

metadata

extra_gated_prompt: >-
  You agree to not use the dataset to commercialize the book contents and make
  commercial OCR software.
extra_gated_fields:
  Institute: text
  Country: country
  I want to use this dataset for:
    type: select
    options:
      - Research
      - Education
  I agree to use this dataset for non-commercial use ONLY: checkbox
license: afl-3.0
language:
  - vi
  - fr
  - en

---

license: cc-by-nc-4.0

This is the dataset in the paper named: "Reference-Based Post-OCR Processing with LLM for Precise Diacritic Text in Historical Document Recognition" (maybe updated later after peer-review process)

The dataset is public for research on high precision Vietnamese OCR, especially for dense text documents like books.

Books in this dataset were published around 1850 to 2003. Please respect the copyright and only used for research purpose! (CC-NC)

• json_raw.7z is zip file for OCR results from Azure Document Intelligence.
• epub.7z is zip file for ebook sources of those files.
• final.zip is zip file of our final pseudo groundtruth.
• page_input.7z is file containing image files of scanned images.
• unlabelled.7z is zip file for all pdf files (of scanned images) without OCR.