|
 |
Building SE Asian Courses - Right Track? | ||
Tags: Study Plan | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| snoonan Triglot Newbie United States learningindones Joined 6252 days ago 23 posts - 56 votes  Speaks: English*, Spanish, Indonesian | Message 33 of 40 19 November 2013 at 3:10am | IP Logged | ||||||||
Seconded -- it is all well received. Wouldn't mind more, actually!
Thanks Leo, this kind of critique helps shoot holes in the weak parts of ideas and helps drive thinking to address them. Thinking about answering your post firmed up some ideas and brought the reading dictionary tools up again (something I had put aside for this first round). I would ask and appreciate more and even harsher criticism. Better now than after we have built everything. :) 1 person has voted this message useful | |||||||||
| Stelle Bilingual Triglot Senior Member Canada tobefluent.com Joined 4145 days ago 949 posts - 1686 votes Speaks: French*, English*, Spanish Studies: Tagalog | Message 34 of 40 20 December 2013 at 11:40am | IP Logged | ||||||||
|
How are the courses coming along? Any news for 2014?
1 person has voted this message useful | |||||||||
| Jeffers Senior Member United Kingdom Joined 4910 days ago 2151 posts - 3960 votes Speaks: English* Studies: Hindi, Ancient Greek, French, Sanskrit, German | Message 35 of 40 20 December 2013 at 12:35pm | IP Logged | ||||||||
When my son was learning to read, he had a lot of the Ladybird Key words reading scheme. The first book starts with a total vocabulary of 16 words, and every 2nd book adds 8 words (if I remember correctly). From that time onwards, I've always felt that a set of readers for adults based on this principle would be ideal for language learners (but with a larger starting vocab, and more than 8 words added with each new story). I agree with Leo that if you keep statistics, this would be simple to do, although it would be very time-consuming and would require a lot of creativity to keep the stories interesting and varied. What you are proposing just changed the whole concept for me, in a good way. You have a set of electronic short stories, the software compares the vocab in the stories to the vocab in your SRS, and tells you how much unknown vocab there would be. Brilliant! Add a feature to automatically generate a list of unknown words, or put them next in your SRS queue, and it would be even better. Of course this wouldn't judge how difficult it is to read in terms of grammar and idiomatic language. But it would still be very useful. 1 person has voted this message useful | |||||||||
| snoonan Triglot Newbie United States learningindones Joined 6252 days ago 23 posts - 56 votes Speaks: English*, Spanish, Indonesian | Message 36 of 40 20 December 2013 at 2:26pm | IP Logged | ||||||||
I should probably put together some running commentary about the process. I just got back from a month in Jakarta where I continued work on the main track, SRS coding and stress tested some of the in-country processes for getting quality voice work and recording. That last part was most important because we can't drop thousands of dollars on sessions that are worthless. As of now, we have a working outline of the main track and have native speakers working through it for Indonesian for just over a month. Writing dialogs and passages -- the first work was poor and boring and we trashed most of it, but it's coming along! We interviewed Indonesian voice contributors and have a good mix. All of this is our test case and process for Tagalog and Malaysian by the spring. You can probably expect lesson material in those languages coming out by late spring. We made some adjustments to how we're writing as well. just started speaking with a few short story fiction authors to adapt and rewrite their work. We started this just yesterday as we had little confidence we'd get capital "I" Interesting writing on our own or with our course writers. It may be a tangent that goes nowhere, but it's worth exploring, I think. So, long story short, Q1 is when Indonesian material should start coming out. Q2 should see the beginnings of Tagalog and Malaysian. We won't have mobile apps until probably July or even later. We hope to start with a rough beta of the SRS (web based, including adaptive audio SRS lessons) probably around maybe April give or take. So, that's where we are now and where we hope to be in the foreseeable future. Please don't get attached to any of these promises slip or change-- sometimes we just have to throw something out and start over if it doesn't work. :)
Yes, yes, yes! This is exactly what we are going for. Every word you read or listen to in a passage gets a touch in your SRS, so they are "added" in that they're bumped into the SRS schedule once you're exposed. That bump may be a light bump if the word isn't very common... That point is new since last month -- the SRS schedule is relative to the word frequency from a massive corpus of real language, so more common words are given priority in the SRS schedule. Second new point, it's not just words, but bigrams and hopefully trigrams & 4-grams if we have the processing power and space to make it work - This means that we can get a (somewhat naive) signal on grammar and usage by knowing how words usually collocate together. Those touch your SRS database too, but I don't know if they will be exposed for interactive study. They will definitely inform scoring of passages, though. So if you have seen idiomatic stuff like "on the other hand" and "hands down" and "hand me that" a few times, that gets picked up in your SRS too. Or at least I hope it will since these databases with 3 & 4 words are getting very large in testing. That may have to come later, but I can say we can still do bigrams at the least. So, to put it together, if you want to read or listen to a text that has the phrase "he jumped up and down" in it, the SRS is doing a lot of things in the background. "he, jumped, up, and, down" is trivial -- you know those words or not? With bigrams, you get 'he jumped' and 'jumped up', which contain implicit grammar and idiomatic usage. trigrams? 'he jumped up', 'up and down'. Getting more idiomatic, natural! 4-grams 'jumped up and down'. So if you are scoring all of that, you are getting a lot of statistical data to score grammar and idiomatic stuff. Not perfect, but still a lot smarter than just words and phrases. Anyway, thanks for recognizing some of the nuance w/ grammar and idiomatic stuff. It means we are solving next problems too, which gets me really excited about getting all of this out there. Edited by snoonan on 20 December 2013 at 2:29pm 4 persons have voted this message useful | |||||||||
| Stelle Bilingual Triglot Senior Member Canada tobefluent.com Joined 4145 days ago 949 posts - 1686 votes Speaks: French*, English*, Spanish Studies: Tagalog | Message 37 of 40 22 December 2013 at 3:16am | IP Logged | ||||||||
|
Thanks for the update! Your approach to SRS sounds *very* interesting!
1 person has voted this message useful | |||||||||
| Stelle Bilingual Triglot Senior Member Canada tobefluent.com Joined 4145 days ago 949 posts - 1686 votes Speaks: French*, English*, Spanish Studies: Tagalog | Message 38 of 40 02 May 2014 at 11:28pm | IP Logged | ||||||||
|
Any news on the programs? Now that it's May, I'm jumping into Tagalog. I'd love another resource, and am thinking
about filipinopod101, but I think that your approach - at least the way that you described it - would work better for me. 1 person has voted this message useful | |||||||||
| snoonan Triglot Newbie United States learningindones Joined 6252 days ago 23 posts - 56 votes Speaks: English*, Spanish, Indonesian | Message 39 of 40 03 May 2014 at 1:58am | IP Logged | ||||||||
I'd definitely say go for whatever you can find out there now. No reason to delay learning! Unfortunately we're still in development and behind the original forecasting on the writing side by months. We've gone through 3 writers which we, of course, didn't plan for. :) I won't go into it, but it's been a challenge keeping writers who consistently deliver on a schedule. Our current writer is doing the work of 2 in half the time and she's been doing great. She's local and we've been keeping it on track with regular in-person meetings which has made a huge difference. But yes, the courses lag behind and depend on this material. I'll post again when we have a clear idea as to when we'll have a course to test. I don't think anyone's actually good at the forecasting game for creative works or software! We'll do what we can though. It's just that we want to have a properly planned and large collection of media to go along with it and we'll keep investing in it until it's good enough to release as a course. Beta here first! One thing that has come up... We're adding a feature and it would be cool to get some thoughts on it. It's designed for intermediate and grow forever with reading skill. A tracking bar with a "study" button. It crawls an L2 page you're reading on demand and scores its readability against your SRS stats, then gives you a word list that you can add to your SRS list. If you do read it and finish, you can rate your understanding and it will impact your SRS. We think it's a nice feature that will bring the reading out of our controlled media and into whatever you prefer to read in L2 online. For example, you could come across a blog post or a wikipedia article in L2 and your SRS would be chugging away behind the scenes, doing its job to let you know how readable it is and what terms you specifically don't know in order to read it. Anyway, that will be in the web reading section and most useful after finishing the main track. 3 persons have voted this message useful | |||||||||
| Crush Tetraglot Senior Member ChinaRegistered users can see my Skype Name Joined 5866 days ago 1622 posts - 2299 votes Speaks: English*, Spanish, Mandarin, Esperanto Studies: Basque | Message 40 of 40 07 May 2014 at 12:32am | IP Logged | ||||||||
|
It's great to hear some news, i've been anxiously waiting to hear something for months (since the last "newsletter"). The web reading bit sounds kinda like Lingocracy's (and i suppose a few other programs') article crawling feature, where you put in a URL and it crawls the page for the article/text then shows you the words you know colored by how well you know them. I think it's a nice feature, it can just be difficult to accurate pull text from every web page, as you often pull extra junk that you don't really need.
As i have zero knowledge of any of the languages you have planned for the immediate future, i'm definitely more interested in the audio courses/other beginner material, but it's neat to see where you see it going as you move out of the beginner and intermediate stages. I'm excited and definitely willing to invest in the course :) 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. |
