Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I’m asking just out of curiosity.

Today I encountered several really strange words while being captchaed by RECAPTCHA:

  • indelms
  • sumbetat
  • polietry
  • grevolfa

If these were reasonable words in any language, a Google search should yield some reasonable pages that use these words in a sentence. However, the numbers of Google results for the above words are 3, 0, 27, and 0. The hits are clearly improbable typos for other, reasonable words.

So where does RECAPTCHA get those words from? (Note: “Books.” is not a sufficient answer :) I’m looking for an explanation for the high incidence of seemingly non-existent words...)

share|improve this question

closed as off topic by Arjan, JNK, harrymc, random Sep 18 '10 at 13:14

Questions on Super User are expected to relate to computer software or computer hardware within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here.If this question can be reworded to fit the rules in the help center, please edit the question.

1  
If this question is off-topic here, which StackExchange site would be more appropriate? –  Timwi Sep 18 '10 at 13:22

1 Answer 1

Well, the answer is books.

They could be scanned wrong, from other languages, or even an author wrote them wrong.

I suggest you read Recaptcha's about page.

Excert

To archive human knowledge and to make information more accessible to the world, multiple projects are currently digitizing physical books that were written before the computer age. The book pages are being photographically scanned, and then transformed into text using "Optical Character Recognition" (OCR). The transformation into text is useful because scanning a book produces images, which are difficult to store on small devices, expensive to download, and cannot be searched. The problem is that OCR is not perfect.

alt text

reCAPTCHA improves the process of digitizing books by sending words that cannot be read by computers to the Web in the form of CAPTCHAs for humans to decipher. More specifically, each word that cannot be read correctly by OCR is placed on an image and used as a CAPTCHA. This is possible because most OCR programs alert you when a word cannot be read correctly.

Edit

As I said, error in OCR -

I believe that Indelms is meant to be Indians - located from United States Statutes at Large Volume 40 Part 2

polietry - most likely scanned wrong from Popliteal - a medical term.

share|improve this answer
2  
@Timwi - updated my answer, again, I am sure it is just OCR errors, if you ever scan a document, you get things like this all the time. It could also be errors in OCR from different languages, I never said it exclusively one of them.... Why would Recaptcha lie about their sources!? –  William Hilsum Sep 18 '10 at 11:29
3  
@Wil: This doesn’t make any sense. You seem to be implying that the words I’m seeing are the result of an OCR. That’s obviously not the case. They are scans from a book. They are scans of those words where OCR failed. –  Timwi Sep 18 '10 at 13:11
2  
@Timwi - Why the attitude? The man is only trying to help (and doing his best as far as one can see). Yes, they are the result of an OCR ... no one said, that OCR is perfect, and that the result is correct. Scan the book, OCR it, verify the result, if there is any doubt as to it, put the original scan as captcha. And yes, they use other languages as well - just the other day I stumbled onto a few words from my own language (very small european language). –  Rook Sep 18 '10 at 13:16
1  
@Rook: Firstly, there is no “attitude” in my comment. I am merely pointing out the flaws with his answer. Secondly: exactly! They put the original scan as captcha, not the result of an OCR (that wouldn’t make any sense). Therefore, “errors in OCR” is not an answer to my question. But these original scans that I’m seeing are nonsensical words from no language at all. If they were words from a reasonable language, even a rare one like Basque or Maltese or whatever, Google would show reasonable pages that actually use the word in a reasonable context. But it doesn’t. –  Timwi Sep 18 '10 at 13:19
2  
@Timwi - No, not necessarily. Google only shows parts of most languages from let's say, last 30 years. Languages change. I have books on my shelf right now with words in them that are no longer used today in neither writing or speech, and will produce very few occurences on google/internet whatsoever. If english wasn't lingua franca of today, it would certanly be a similar case (Try comparing Shakespearean english and American english ...) –  Rook Sep 18 '10 at 13:38

Not the answer you're looking for? Browse other questions tagged or ask your own question.