- Home
- Searching scanned Arabic archives
How to make a scanned Arabic archive searchable
Most organisations in Oman hold years of letters, decisions and reports that exist only as scans or paper. Nobody can search them, so they are rarely used. With the right preparation they can answer questions like any digital document.
Start with an inventory
- Where is it? Shared drives, document systems, email attachments and boxes of paper.
- What matters most? Pick the collections people ask about every week, such as decisions, contracts or correspondence with a key partner, and start there.
- Who may see it? Agree access rules before anything becomes searchable, so the archive does not reveal more than it should.
What makes a scan readable
- Scan at a resolution of at least 300 dpi. Lower settings blur Arabic letters and dots.
- Keep pages straight and the whole page in frame, including stamps and margins.
- Prefer clean greyscale or colour over very dark or high-contrast black and white, which can erase faint text.
- Keep the original file. Searchable text is added alongside it, never instead of it.
How Falaj uses the archive
Falaj reads scanned and photographed documents in Arabic and English, adds them to the same library as your digital files, and marks answers that come from scans so people know to glance at the original. Staff then ask in their own words, in either language, and get an answer with the document and passage it came from. Everything runs on your own server in Oman.
Common questions
Can Arabic scans really be searched?
Yes, when the scan quality is reasonable. Falaj reads printed Arabic and English from scanned and photographed pages and makes them searchable with your other files.
Do we need to rescan everything first?
Usually not. Start with the existing scans, see which ones read poorly, and rescan only those.
Does the archive have to be moved?
No. Falaj reads from where the files already are, and the originals stay untouched.
See it on documents like yours
We bring Falaj to your offices and run it on a sample of your own files, so you can judge the answers yourself.