Anonymous ID: 5c6810 Dec. 11, 2019, 1:47 p.m. No.7483223   🗄️.is 🔗kun   >>3271 >>3284 >>3304 >>3317 >>3343 >>3507 >>3518 >>3561

>>7483103

>>7483124

 

Whether this is deliberate obfuscation is up for debate but what this boils down to is document search. Optical Character Recognition takes a document that is typically an image and makes it into searchable text. To do this an OCR engine, typically Tesseract and a proprietary algorithm made by whomever (ABBY, etc.) then parses the document to identify letters. This is how a document becomes searchable.

 

Resolution and quality of the scan is YUGE and will exponentially effect the search ability of a document. Notice how cl looks like d when I type ( c l ~ d ). If the OCR engine (or someone deliberately tampering with the document) identifies a c l as a 'd' and a user is searching for whole words only and spells out Clinton, the document will return no results because of that errant 'd'. This COULD be a way that documents were hidden in archives, or IT COULD just be temperamental OCR and government workers that don't know any better, so it is debatable.

 

What you see in anon's post are search terms for documents that net results. Instead of searching for Biden, look for letters or numbers that could be mistaken for an i ( 1, L, etc) or a d ( c l ) etc. This is what you are looking at in anons post: anon did the work and apparently has had luck with search terms.

 

The important thing is the matching - you'll hear some go deep into the 'this is how they hid the documents' and others like me taking the 'We've worked with OCR, the technology is temperamental at best'. Perhaps it is a bit of both but the thing to understand is now we have a document searching strategy. Hope this helps your Document Fagging.

Anonymous ID: 5c6810 Dec. 11, 2019, 2:03 p.m. No.7483345   🗄️.is 🔗kun   >>3422

>>7483284

>>7483271

 

I'm catching up on today's posts anons. I'll say that I am trending toward fuckery after all the discussion.

 

>>7482552 PB

 

I do not have enough software fagging experience to know what is needed to deliberately change the names (light script any script kiddy could run to find + replace whatever). Willing to bet there is some sort of trail or forensic evidence somewhere of this fuckery. Maybe.

 

I'm off the OCR slide - only reason I brought it up was anon asked. I'm with you both though after reading a few PB posts. Onward.