Live data from Hacker News

Optimising the T9 Keyboard

torvaney.github.io

11–17 of 17 posts

Re: Optimising the T9 Keyboard

#11
This one is fun. Maybe one optimisation worth checking is what's possible if you have to stay alphabetical.

E.g. T9 puts 4 letters on '9', presumably a good choice as 'X' and 'Y' are rare. Would it be worth it to shift the 'V' from '8' to '9'. '3' has 'D' and 'E' on the same key, maybe move the 'D' to '2' ?

Re: Optimising the T9 Keyboard

#12
post #6

I don't think this is very insightful. Using the first half of books for training data and the second half for testing data is still training the model specifically for these texts and authors. Not quite as bad as testing on the training data, but not great.

Hello! OP, here. I agree that the task is training for this specific subset of English writing, which isn't ideal.

For this task, I was primarily interested in whether the task would work at all. My assumption is that given we can optimise for these texts, we could optimise for more representative datasets, too. Perhaps you think this is a weak assumption?

Do you think testing on a sample of totally different texts from different authors would be more convincing?

Re: Optimising the T9 Keyboard

#13

This is great! I like your site and would like to continue getting updates from it. Would you consider adding an RSS feed?

I tend to blog more often on my soccer blog (https://www.statsandsnakeoil.com), which has RSS set up. Naturally, that's a bit more niche. I'll look into RSS for this site if I start adding more new posts and projects.

Re: Optimising the T9 Keyboard

#14

This one is fun. Maybe one optimisation worth checking is what's possible if you have to stay alphabetical. E.g. T9 puts 4 letters on '9', presumably a good choice as 'X' and 'Y' are rare. Would it be worth it to shift the 'V' from '8' to '9'. '3' has 'D' and 'E' on the same key, maybe move the 'D' to '2' ?

Yes, it's interesting that the T9 seems to have made a slight improvement on my "naive" alphabetical layout. You might even be able to brute-force a solution where the alphabetical constraint is imposed.

Re: Optimising the T9 Keyboard

#15
post #4

Pretty cool read. My biggest problem with T9word is Spelling. If you misspell something on a full keyboard, the other person may be able to ges by sounding it out. T9word would throw out a word that would be no where near what you want. 'Ges' would become 'her' and the other user would never think 'guess' was the intended word. My 2nd biggest problem with T9word is you need to look at the screen to verify the word yo…

Absolutely. This will be an even bigger issue for people texting in a second or third language (which I understand is quite common among countries and communities where feature phone use is high).

As I understand it, one of the big advantages of T9 had over other, more sophisticated forms of predictive text is that each word in the dictionary can be encoded in close to 1 byte. Given the constraints faced at the time, T9 feels to me like it is close to a local optimimum.

Re: Optimising the T9 Keyboard

#16
post #2

Awesome research! I was just thinking the other day, about how a qwerty layout for T9 might have been easier and more efficient. Cool to see something even better. A little off-topic, but this 6-button keyboard format looks perfect for my dream device. I take a lot of walks. I often listen to an audio book while I do so. I also frequently get texts, or hear something in the book I may want to write down. I've always…

I remember back when I could ride a bike while texting and still watch the road. T9 was the high point of text input on handheld devices!

Re: Optimising the T9 Keyboard

#17
From what I can remember from using a t9 keyboard quite a long time ago, was that you pressed each key to cycle through the letters on the key. with a break if you need multiple letters on the same key after each other.

the whole predictive word thing was something pretty much alien to me.

Post reply on HN