Title Finder Program Results
This topic contains 2 replies, has 2 voices, and was last updated by Cain 2 years, 4 months ago.
February 10, 2016 at 3:33 am #6217
One of the things I’ve been focused on is finding valid titles on the shelves. I figure the chances of this are much more likely than finding a page like you can get with the random english words search tool.
I’ve done some scraping of the site (sorry Jonathan), looking at the first page and title of over a hundred hexes (or 64,000 books). I then ignore the periods and commas, in the title and see if any words show up. Also, since I needed to request the first page to get the title, I also count the number of characters on the page.
I’ve only found a couple of books with a legible title.
scan gym – hex: yetanotherexperimentrat14 wall 1, shelf 2 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat14-w1-s2-v28:1
,bye – hex: yetanotherexperimentrat30 wall 1, shelf 2 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat30-w1-s2-v08:1
jzfcpf hand of.h goze – hex: yetanotherexperimentrat41 wall 3, shelf 5 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat41-w3-s5-v05:1
a bum – hex: yetanotherexperimentrat57 wall 3, shelf 4 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat57-w3-s4-v14:1
The longest distinct words in a title that I’ve found is 5 letters, wishy and chevy, the two books located:
mztehr wishy.kotth – hex: yetanotherexperimentrat31 wall 3, shelf 1 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat31-w3-s1-v10:1
.chevy,vm,dxywnh,gbg – hex yetanotherexperimentrat49 wall 4, shelf 1 – https://libraryofbabel.info/book.cgi?yetanotherexperimentrat49-w4-s1-v06:1
In terms of character count, I have found it to easily within one standard deviation from the expected norms per character, after a couple million characters counted. This would be pretty consistent with what would be expected from an algorithm that uses psuedu-random data to populate the books, so I wasn’t really surprised by it.
One thing I would like Jonathan Basile to add to the site is something that would make the task of finding legible titles much easier. A way to pull a whole hex or walls worth of titles in one request. I would feel less guilty doing 1 page request to get a walls worth of data rather than 160.February 10, 2016 at 7:56 pm #6223
I’d like to work on a sort of API that would made requests like these easier – it’s one of many things that I need to find the time for. Hopefully in the not-too-distant future!
Good finds! I especially like scan gym.March 23, 2016 at 2:33 am #6479
I feel like The Hand of H. Goze would be a great title for a book.