|
 |
Super-fast vocabulary learning techniques | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| luhmann Senior Member Brazil Joined 5330 days ago 156 posts - 271 votes  Speaks: Portuguese* Studies: Mandarin, French, English, Italian, Spanish, Persian, Arabic (classical) | Message 201 of 255 20 April 2010 at 3:26pm | IP Logged | ||
|
I am currently swallowing a Chinese dictionary so fast I can't believe.
I use a SRS I wrote myself that manages scheduling up to the second. Sucessfuly tested cards will be shown again at ever doubling intervals, for instance 30sec, 1min, 2min, etc. Failed cards are usually just reset as if they were new. At first I had words in order of frequency, but soon I realized that learning words with known characters was very easy, whereas words with new ones were very hard. So sorted the lexicon in logical way, rather than by frequency. I created a list of characters by picking a radical, then adding all characters made only of known radicals, then I pick another radical, add more characters, and so on. Once I sorted the character list, I applied the same logic to sort the lexicon (some 30,000 words I compiled from various sources) in a way that I first learn every existent permutation of known characters before learning the next. This way cards have a great degree of mutual reinforcement. I started using this method about two months ago, and I have added on average 75 new words per day. But as I learn more characters and there are more possible permutations, it keeps getting easier and easier. Last weekend I got carried away and added over 400 new words. PS Don't ask me for the program because the source code is really messy and needs to be rewritten from scratch, it is hard-coded to my data and it cannot be used without understanding its innerworkings. 4 persons have voted this message useful | |||
| tommus Senior Member CanadaRegistered users can see my Skype Name Joined 5863 days ago 979 posts - 1688 votes Speaks: English* Studies: Dutch, French, Esperanto, German, Spanish | Message 202 of 255 20 April 2010 at 5:18pm | IP Logged | ||
What do you mean exactly by radical? Is that somewhat specific to Chinese, or is it what might also be called root? I think learning all related words, (possible also all similar-looking words even if they are not related) as a single group would be very productive. I have tried this on small examples, but your suggestion may inspire me to do it on my main second-language dictionary. I have what is probably a similar software environment to yours, where I mess around with dictionaries and different ways to view them. So I'll look at some new ways to look at related words in groups. Thanks for the inspiration. 1 person has voted this message useful | |||
| brian91 Senior Member Ireland Joined 5441 days ago 335 posts - 437 votes Speaks: English* Studies: French | Message 203 of 255 20 April 2010 at 6:31pm | IP Logged | ||
|
This is an awesome thread, and I wish I discovered it sooner. Bookmarked for study after
my German oral exam tomorrow. :D I'm trying to learn the three thousand most common words in German at the moment and I hope it will really help my conversational skills. 1 person has voted this message useful | |||
| luhmann Senior Member Brazil Joined 5330 days ago 156 posts - 271 votes Speaks: Portuguese* Studies: Mandarin, French, English, Italian, Spanish, Persian, Arabic (classical) | Message 204 of 255 20 April 2010 at 6:33pm | IP Logged | ||
Radicals are the basic components of Chinese characters ( see http://www.zhongwen.com/ ). They are not akin to the root of words, characters are completely automomous lexical entities, its components have scarcely any bearing on meaning. The root of the Chinese words would be the characters. Edited by luhmann on 20 April 2010 at 6:35pm 1 person has voted this message useful | |||
| michau Tetraglot Groupie Norway lang-8.com/member/49 Joined 6223 days ago 86 posts - 135 votes Speaks: Polish*, English, NorwegianC1, Mandarin Studies: Spanish, Sign Language Studies: Burmese, Toki Pona, Greenlandic | Message 205 of 255 20 April 2010 at 11:23pm | IP Logged | ||
Your method sounds great. I have a few questions: - how much time (on average) do you spend on repetitions and adding new material every day? - what is your forgetting index? i.e. what is the percentage of items you do not remember if you take a random sample of the items you've already learnt? - I understand you don't want to share your code, but how about just the word database, sorted according to the rules you mentioned? It would be extremely helpful. I would like to compare effectiveness of your method to mine. I'm also learning Chinese characters using SRS (Supermemo for Pocket PC). I use a database of HSK A and B words, sorted according to frequency. The forgetting index is fixed at 10%, I believe (at least if my Supermemo has the same index as Supermemo for Palm). I add 16 words per day on average, and spend about an hour per day on learning and repetition. 1 person has voted this message useful | |||
| luhmann Senior Member Brazil Joined 5330 days ago 156 posts - 271 votes Speaks: Portuguese* Studies: Mandarin, French, English, Italian, Spanish, Persian, Arabic (classical) | Message 206 of 255 21 April 2010 at 3:13pm | IP Logged | ||
I am very irregular, I normally do long sessions on weekends, probably over 4 hours over the day. During the week I try to find some time to review them during my workhours, but I find hard to concentrate as there are lots of people talking near me. I only make the point to test the cards with short intervals, the backlog keeps growing till I can find some time. The forgetting index also very variable, When I run into a long sequence of words having one common character, the tendency is that I will recall flawlessly, whereas if a character has few example words, I keep forgetting it often. You will find table of character component elements, sorted (approximately) in the way I have delineated here: http://freetexthost.com/nudmgxgnbk . The data on character components have many flaws though, I have created the table by parsing the html files from http://zhongwen.com , but not all files were parsed correctly. As for the sorted wordlist, I'll try to find it later. 1 person has voted this message useful | |||
| puffin1981 Diglot Newbie Hungary cookingstar.blogspot Joined 5308 days ago 4 posts - 4 votes Speaks: Hungarian*, EnglishC2 Studies: Italian, Spanish, Thai | Message 207 of 255 13 May 2010 at 12:49pm | IP Logged | ||
|
Hey guys, I work as an English teacher and teacher trainer and the website we keep
recommending to people is http://www.quizlet.com Possibly the best (vocab) learning site ever. After registration, which is free by the way, you can create your own flashcards, or use sets that others have made. Then you can choose to familiarize yourself with the cards, learn the cards, test yourself or play games with whatever you would like to learn. I haven't read all the posts in this thread yet, so if someone has already mentioned quizlet, apologies... otherwise go and give it a go :) 1 person has voted this message useful | |||
| luhmann Senior Member Brazil Joined 5330 days ago 156 posts - 271 votes Speaks: Portuguese* Studies: Mandarin, French, English, Italian, Spanish, Persian, Arabic (classical) | Message 208 of 255 13 May 2010 at 2:45pm | IP Logged | ||
|
Sorry I could not find anywhere I could post my wordlist (a 300k long utf-8 file), here is an extract from it that should illustrate what it looks like: http://freetexthost.com/5aeqpl1py2 .
I could send the full it by e-mail to anyone interested. However, its usefulness would be limited, because it does not strictly follows the principles I have delineated in my previous post -- it was a slow development, I first started in an assistematic way, then started looking for ways to improve it. The major problem of my wordlist, as it is right now, is that sometimes characters that have little or no example words are added. My next step would be to identify those, and postpone their apparition until enough words can be built. Another thing I will do when I have the time is to add example sentences too. I think this kind of sorting would work very well in conjunction with the 100k sentences method. I do plan making the next version system available, but I will only have time to work on it on my vacations. Edited by luhmann on 13 May 2010 at 3:20pm 1 person has voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
