Including names and rare obscure words with various usages, well over 600,000 words in all.

This took almost 2 months, with one computer going all night, all day... And it just finished today!!!

It could have been done much faster, but Google doesn't like me doing this, so they would block me frequently.

I now have a massive dictionary file, with ratings based on hits returned for every English word imaginable.

I am now making a program to give me more statistics from the data I've collected.

For example, it estimated that Google has currently cached 2,228,694,630,626 English words in total. It has only found 33 instances of words that are exactly 33 characters in length, but has found 170,351 instances of 34 character words. It is also safe to assume that 4 letter words encompass the bulk of the current English language with over 371,163,887,436 instances.

Why am I doing this, you might ask? I'm working on a program that infuses keyboard usage with a simple, but highly accurate AI. It will finish all your words for you, correct spelling mistakes, and allow you to type with one hand if you wish (even faster than you already do with both hands), without the need for modifier keys or a steep learning curve. It will also learn how you type, by what words you choose, and thus increase it's accuracy over time. In essence, it will make typing a much faster and easier task. There is software "like" this, but nothing nearly as accurate or as easy to use as what I'm working on.

I have alot of optimization to do, as currently the program takes over 14 seconds to load, and it can take over 160ms for it to compute all permutations of 10 character words (which is unacceptable)... But this is already 10 times faster than my first version of the software...

When I finish the first beta, I'd like to get some testers from the crew, if possible.

Dibs on the Beta!

what do you plan to do with it when you finish, and how long to estimate it will take you to complete?

Holy crap, that's some intense shit right there.
 And how will the program predict the words? Will it look at the first 2-3 letters and then determine what is the most likely word choice based on previous words and frequency of word use? Or did I totally misunderstand everything.
It sounds cool regardless.

Quote from: SirCannabis;1006699what do you plan to do with it when you finish, and how long to estimate it will take you to complete?

I plan to give it away. There is no (decent) typing assistant software that is free, or even open-source. This has numerous applications, including those who are handicapped, or are fighting carpel tunnel. It's my hope that if this is freeware, that it'll spread quickly and more people will get it.

Of course, I will put it on a site with a bunch of ads, and will make frequent releases, and hopefully that will drive enough traffic to make a buck off of it.

I may also embed a kind of Google search bar (with a hotkey), so that I could earn referral points when a user searches with it.

did you googled franfurr?

Quote from: Behemoth;1006704And how will the program predict the words? Will it look at the first 2-3 letters and then determine what is the most likely word choice based on previous words and frequency of word use? Or did I totally misunderstand everything. It sounds cool regardless.

Yes, basically... it does a bunch of math based on the most used words and the most used lengths based on what you have already typed, depending on the mode you are using.

How about common sentences and phrases. Can your AI be programed to complete sentences and/ or phrases that its user types often?

Quote from: Mr. Blue;1006708I plan to give it away. There is no (decent) typing assistant software that is free, or even open-source. This has numerous applications, including those who are handicapped, or are fighting carpel tunnel. It's my hope that if this is freeware, that it'll spread quickly and more people will get it.

Of course, I will put it on a site with a bunch of ads, and will make frequent releases, and hopefully that will drive enough traffic to make a buck off of it.

I may also embed a kind of Google search bar (with a hotkey), so that I could earn referral points when a user searches with it.

would probably look great on a resume as well

dibs on beta as well

Quote from: Anchor;1006724How about common sentences and phrases. Can your AI be programed to complete sentences and/ or phrases that its user types often?

I haven't really begun to plan that phase, but I do intend on eventually putting a full language analyzer into it, so that it can predict common phrases, and also correct grammar... BUT I gotta start with the baby steps...

Are you a genius? I'm not being sarcastic, either. This is a really intuitive program.

These kinds of accessories are why I do all my typing in notepad.

Neat statistics though!

This sounds really cool. I'm always spelling definitely wrong, it gets frustrating.

Sounds interesting, but what is this caching, webpages or webchats? Dialogue and use of the english language varies a great deal in those circumstances.

Sounds incredibly useful though, I look forward to it.

I third Calling Dibs on beta!

i call ultra dibs on alpha, because im 1337

now google every word of china

Quote from: RupeeClock;1007004Sounds interesting, but what is this caching, webpages or webchats? Dialogue and use of the english language varies a great deal in those circumstances.

It would only cache single words, and only if they are out of the usual statistics (something unique to your typing style).

But you are right, there are huge differences between the typical typing style between such programs, perhaps I'll have the caching differentiate between word processing and chat.

I'd like being a tester, it would help my homework process most likely if it worked.