Observed arrival · 2026-10-06
Nassakh’s page-linked Arabic book workflow
Nassakh presents a workflow for turning scanned Arabic books into reviewed text linked to images of the printed pages, with tools for checking quotations and producing new editions.
- For
- Arabic book publishers, editors, and researchers
- Worth noticing
- Unresolved OCR differences are routed to “needs image match,” rather than counted as a misquotation.
Field notes
The homepage lays out six stages, from separating text, footnotes, and page numbers to producing a new edition. It shows two Arabic-reading models compared with Tesseract and a disputed word linked to its printed page image. The citation checker distinguishes a text difference from an unresolved reading and from a passage absent in the account’s available books; the page says account books are visible only to members.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue