|
 |
Designing a language learning app | ||
Tags: Software | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| Akalabeth Groupie Canada Joined 5505 days ago 83 posts - 112 votes  Speaks: English* Studies: German, Japanese | Message 1 of 4 17 December 2011 at 8:19pm | IP Logged |
|
So I'm designing a language learning program somewhat similar to Anki, and am curious
if anyone has any ideas re: design issues. Long story short, I just got an iTouch, and found the MindSnacks French app, which I really like (in particular the Fish Bowl game). There are some downsides to it (you can't use custom words for one), so I'm trying to make a similar desktop version of it. The basic idea is you would get given a word in English (or maybe you're target language), and then have to choose the translation by multiple choice. You have a limited amount of time to answer, which decreases with each successful answer, and after a certain number of failures the game ends. The thing I really liked about the iTouch app was that there's no indication if you have more reviews to do, or if you're reviewing early, which has made it seem less of a grind than Anki. You just review as much as you like, and every once and a while it presents new words. It seems to use SRS still, but in a weird way: it will present 20 new words, and you'll mostly only study those, but it will occasionally add old words into the mix. Minimally, what I want to make is something game-y to be a little less of a grind, to accomplish fast recall and overlearning. Things that other people's ideas might help with: 1) SRS algorithm. The Anki algorithm requires you to grade yourself every question, and I'm not sure how well it works if you frequently review early, so I need to decide on something different. 2) How the questions are presented. Is multiple choice the best way to do fast reviews? If multiple choice, how many answers, and how are the other choices chosen? 3) How much time to give for each question. I'm thinking it could slowly increase/decrease the amount of time you are given depending on how well you do, to try and find your limits. Any other ideas people have are welcome, but I'm not really a great programmer, so I'm focusing on functionality over features. I'm intending to use this as a supplement for Anki. 1 person has voted this message useful | |
| Hampie Diglot Senior Member Sweden Joined 6645 days ago 625 posts - 1009 votes Speaks: Swedish*, English Studies: Latin, German, Mandarin | Message 2 of 4 17 December 2011 at 8:27pm | IP Logged |
|
Multiple choice is bad, it presents you the answer in a context that you normally would not have. You will ned get
to choose from a list of similar words when you encounter a word in reality, but only the word where it is. 1 person has voted this message useful | |
| Akalabeth Groupie Canada Joined 5505 days ago 83 posts - 112 votes Speaks: English* Studies: German, Japanese | Message 3 of 4 17 December 2011 at 8:44pm | IP Logged |
|
I think the main reason I like multiple choice is you can quiz yourself the fastest. I'm
curious what the effects of drilling yourself for near-instantaneous recall would be. But yeah, multiple choice is a pretty non-natural way to study I think. In order to do the timed reviews the program will need to be able to evaluate whether you got the correct answer itself, and "Did you click on the correct button" is definitely the simplest way to do it. You could also type in the answer, but then that has the issue that there might be multiple valid answers, plus typos. 1 person has voted this message useful | |
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 5997 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 4 of 4 17 December 2011 at 9:13pm | IP Logged |
|
The one hint I'd give you is that even a guess has a 25% chance of being right, so your algorithm shouldn't automatically tag every correct answer as good, and automatically push the review time back -- there needs to be some kind of averaging of results to measure that the item is known.
1 person has voted this message useful |
If you wish to post a reply to this topic you must first login. If you are not already registered you must first register
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
