31–40 of 95 posts
Fairly frequently, OCR engines are posted here. But almost without exception, they lack layout analysis, which renders them largely useless. Is this something that could be combined with those OCR engines? (e.g. TesseractOCR...)
who would be interested by an online website doing the job?
if you really want to rake it in, serve, at static speeds (meaning instantly, I swear, boot a ramdrive (Tmpfs) and serve static html from nginx all from RAM), text versions of the top 10,000 web sites. there is so much crap on most sites. re-crawl hourly. monetize via Google adwords. EDIT: I'm not sure why I'm being downvoted. I am not suggesting serving PDF's. I am suggesting serving tiny text renders of top sites,…
Fairly frequently, OCR engines are posted here. But almost without exception, they lack layout analysis, which renders them largely useless. Is this something that could be combined with those OCR engines? (e.g. TesseractOCR...)
Earlier quoted context omitted.
if you really want to rake it in, serve, at static speeds (meaning instantly, I swear, boot a ramdrive (Tmpfs) and serve static html from nginx all from RAM), text versions of the top 10,000 web sites. there is so much crap on most sites. re-crawl hourly. monetize via Google adwords. EDIT: I'm not sure why I'm being downvoted. I am not suggesting serving PDF's. I am suggesting serving tiny text renders of top sites,…
You can use the SHA-1 of the PDF's to avoid serving the same pdf twice.
who would be interested by an online website doing the job?
See e.g. my file sharing app https://github.com/andreif/SecretFile
Fairly frequently, OCR engines are posted here. But almost without exception, they lack layout analysis, which renders them largely useless. Is this something that could be combined with those OCR engines? (e.g. TesseractOCR...)
I would not call these services useless ;) - but I wonder the same... Some apis like https://ocr.space return the coordinates of each converted word. Can that be a used input? (I have not tried it yet)
Although I haven't tested this yet, these utilities tend to fail when fed a table with empty cells.
For those interested in converting PDF tables into CSV, there's also Tabula ( http://tabula.technology/ ) (Used by many journalists to analyze the data in PDFs)
I find it absolutely ridiculous that we have to resort to these kinds of tools :/ We have digital formats, and we decided to standardize document distribution on the one that makes it as hard to extract data as if it were on physical paper.
PDF is the successor of PostScript. PostScript is a stack-based programming language where anything can happen, while PDF enforces some document structure and metadata structure on top of it, so you can e.g. at least determine where pagebreaks are, without having to interpret ("run the code of") the whole document.
Still, PDF is simpler than PostScript in the same sense that XML is a simplification of SGML. Jumping from PDF to a well-designed format would be like jumping from XML to JSON or S-Expr.
Earlier quoted context omitted.
if you really want to rake it in, serve, at static speeds (meaning instantly, I swear, boot a ramdrive (Tmpfs) and serve static html from nginx all from RAM), text versions of the top 10,000 web sites. there is so much crap on most sites. re-crawl hourly. monetize via Google adwords. EDIT: I'm not sure why I'm being downvoted. I am not suggesting serving PDF's. I am suggesting serving tiny text renders of top sites,…
Doesn't Opera Mini or Turbo already provide this sevice? Perhaps add PPMD proxy text compression with an English dictionary with a JavaScript browser plugin on top of that. You can't get more efficient than that