OCR_Task_JA / README.md
xiangr's picture
Update README.md
0829a6c verified
|
raw
history blame
2.21 kB
metadata
license: cc
language:
  - ja

Japanese_OCR

This dataset contains four collections of Japanese government and financial institution PDF documents, intended for OCR research, document layout parsing, and financial text extraction.

πŸ“‚ Contents

Folder Description Source Institution Size
FSA_Whitepaper Annual white papers published by the Financial Services Agency (ι‡‘θžεΊ) FSA 1.5 GB
BOJ_PDFs Research papers and discussion papers from the Bank of Japan (ζ—₯ζœ¬ιŠ€θ‘Œ) BOJ 464 MB
JPX_PDFs Monthly equity market reports from Japan Exchange Group (JPX) JPX 450 MB
SESC_Report Annual activity reports from the Securities and Exchange Surveillance Commission (SESC) FSA - SESC 577 MB

Total size: ~3 GB

🧾 Licensing and Usage

All documents are publicly available from the official websites of Japanese financial and regulatory authorities. However:

  • JPX PDFs are labeled "軒載禁歒 (no redistribution)" β€” included here solely for research and non-commercial OCR use.
  • BOJ PDFs are research papers generally available for non-commercial academic use.
  • FSA Whitepapers and SESC Reports are government publications intended for public information and transparency.

Please verify licensing status if you intend to use these documents for redistribution or commercial purposes.

πŸ” Use Cases

This collection can be used for:

  • Optical Character Recognition (OCR) benchmarking
  • Document layout analysis (tables, headings, multi-column)
  • Japanese financial terminology extraction
  • Training document parsing models in multilingual or financial settings

🏷 Tags

OCR, Japanese, PDF, financial-documents, government-data, non-commercial, public-sector