Shadeless Wrote:

vinniram Wrote:mate, you're awesome I did exactly what you said and it works like a charm Thanks a million!!! I think I can finally stop mucking around getting kore to work, and actually start *using* it now

I'm glad to hear that Good luck!

Pfft. n00bs. Up up down down left right left right AB select start worked for me.

nest0r Wrote:

Shadeless Wrote:

vinniram Wrote:mate, you're awesome I did exactly what you said and it works like a charm Thanks a million!!! I think I can finally stop mucking around getting kore to work, and actually start *using* it now

I'm glad to hear that Good luck!

Pfft. n00bs. Up up down down left right left right AB select start worked for me.

It never occurred to me that I could use gamepad for reviewing, you brought meaning of "Playing with Anki" to whole new level

nest0r Wrote:

Shadeless Wrote:

vinniram Wrote:mate, you're awesome I did exactly what you said and it works like a charm Thanks a million!!! I think I can finally stop mucking around getting kore to work, and actually start *using* it now

I'm glad to hear that Good luck!

Pfft. n00bs. Up up down down left right left right AB select start worked for me.

Weird, I had to use ↑ ↑ ↓ ↓ ← → ← → B A

http://en.wikipedia.org/wiki/Konami_Code

cangy Wrote:I haven't had a chance to move anything yet, but I guess you can get the attachments from the above link for now...

now here: https://sites.google.com/site/ankinihongo/

I thought this might help more kore learners: since the words in kore come in bold in the sentences by default, you can tweak that to make it even more clear (in my case, I preferred KO2001's blue). You only need to add a style declaration to your cards:

Code:

<style type="text/css">[style="font-weight:600;"] {color:blue;}</style>

And your card could look like this (NB The Japanese sentence without furigana is what's visible at first, then the requested word and meaning of word and sentence are added on the answer side). I prefer to see the words in context, so I don't have to produce both [kana]dondon[/kana]teadily and [kana]dondon[/kana]:drumming_sound at the same time and this is a possible way to do that.

Futhermore, I also added listening cards. Very nice that this is possible with kore. I hope to find out soon whether that helps/works. The point was I had the impression that I was taking too much info out of the kanji. Listening-only forces you to recognise a (spoken) word just by the phonemes. I hope this'll also reduce the amount of failures I have on some kana-only cards, since they'll come by twice as often now. Of course the context component isn't there, but hearing an entire sentence just takes too long.

Lastly, I started using swanki, but I'll report on that in a different thread.

okay this is going to sound really stupid, but how come the Kore deck has 6000 cards if it is core2k+core6k; shouldn't it be 8000 cards?

Core 2k is 2000. Core 6k is 4000, bringing the total to 6000

They're tricky like that.

KanjiDevourer Wrote:I thought this might help more kore learners: since the words in kore come in bold in the sentences by default, you can tweak that to make it even more clear (in my case, I preferred KO2001's blue). You only need to add a style declaration to your cards

nice! actually, that gives me an idea...

now with cloze deletion on vocab in sentences

https://sites.google.com/site/ankinihongo/

cangy Wrote:now with cloze deletion on vocab in sentences

https://sites.google.com/site/ankinihongo/

That's pretty cool. Though it's not worth it for me doing now, had I had this earlier I'd do a find/replace of "<b>( )</b>" with "<b>[...]</b>" to keep it in line with Anki. Probably a good choice for those that like to do Kana to Kanji type cards

Q: Kana Word -- Clozed Sentence
A: Kanji Word -- Kanji Sentence

or for a bigger challenge

Sadly, Kore doesn't give enough info for fill in the blank without some type of trigger. The photo is only for 2k and even that's not enough a lot of the time.

Nukemarine Wrote:I'd do a find/replace of "<b>( )</b>" with "<b>[...]</b>" to keep it in line with Anki.

I tried that at first but it conflicts with the furigana markup and this is more in line with standard Japanese texts, though I might change it later if I get round to improving xxfurigana...

Nukemarine Wrote:Kore doesn't give enough info for fill in the blank without some type of trigger.

You could use the English vocab and sentence too but I guess that changes the nature of the exercise?

Ah, yeah, I noticed there was furigana problems when you try to use color coding. Guess it applies to elipses as well for some reason.

About triggers: It's kind of a give and take on whatever one decides. With fill in the blank, one can use images, context, or another language to determine the user pretty much can create the word on the fly. With Kana or Kanji to Kanji or Kana type cards, I guess it's just an improvement of reading ability.

On one hand, it's great to have the ability to inately know the word for whatever concept. On the other, such production type cards take up a lot of time. Guess one can try (or be led to do by an instructor) to depend on production cards for beginner and basic words. After that, move on to just recognition cards for the larger subset of intermediate cards.

The cool thing about all this is the user (or tutor) can set up their program as they see fit and adapt/change as time and ability goes on.

What sort order would make sense if i was only into the vocab (single words)? Or wouldnt it make sense at all with this deck?

Tanhuser Wrote:What sort order would make sense if i was only into the vocab (single words)? Or wouldnt it make sense at all with this deck?

The default (iirc, rtklite+ko2001) works reasonably well for just using the deck as a vocab list (which is what I do, although I read the sentence when learning a word for the first time to get a better idea of the meaning). The downside is that kana only words are somewhat frontloaded, but you can just suspend kana only words and slowly add them when you feel like it (I do this).

It probably could be better, but the order seems good enough after the first few hundred words.

mgbp7 Wrote:Okay, I finally have kore up and running, audio and all, in Anki. Did my first reviews yesterday.

So anyway, I hope to be able to come back after I've made some progress and give some feedback on what kind of impact studying the deck has made on my reading ability.

I said I’d give an update at some point on my progress with the Kore deck, so…. First of all, my goal initially for Core 6000 was to finish by the end of this year at a pace of 20 new words per day (which really comes to 40 per day as I’m doing forward/reverse vocab cards, i.e. kanji – kana as well as kana – kanji).

I finished RTK last year and figured that Core 6000 would be a nice follow-up and would give me a base of common vocab/kanji compounds. My hope is to use that basic vocab to focus as of next year on slogging my way through native Japanese material -- bilingual and children’s manga, online sites + rikaichan, etc. I’ll probably be an idiot and also import FF Versus XIII from Japan as I did with FFXIII. (This assuming the game comes out in 2012, which isn’t a given.)

I have about 1 hour a day for Japanese study, so this essentially means that my focus this year is Core 6k. I squeeze in a bit of Genki II study during lunch at work and RTK review (I plan to keep reviewing indefinitely as my Japanese reading ability is way too low to allow for exposure to enough kanji to merit dropping reviews). I also try to look at native Japanese text (Yahoo! Japan Kids, etc.) when I can and listen to Japanese audio regularly.

Anyway, I’m currently about ~1200 into the Core list. I’m studying by JLPT level –-started with 4, went through 3, and am now on JLPT 2 level words. I am definitely seeing the words I’m studying in Core out in the “real world”. A lot of times I will see a given word within a day or two of having learned it in Core. I can usually recognize it immediately (reading and meaning). So this is encouraging. I’ve even noticed a difference with the smattering of kanji posted on this forum –- I’ll lazily glance over it not expecting to understand it and then when I look more closely I’ll realize I can actually read it. Pretty nice. Even when listening to audio I have found myself recognizing Core vocab.

And really this is the only reason why I’ve been able to stick with Core, because I find having to review 100+ words and then drill 20 new words on top of it on a daily basis tiring, personally. Although I aim for 1 hour of Japanese study a day after work, up until recently it was usually taking more like 1.5 – 2 hours. I also found myself questioning the wisdom of drilling random vocab cards -- kaikei/kankei/keiken/shuchou/koshou/kouchou start to all run into each other and don’t stick too well long term.

I made some adjustments, which has cut my Anki review/study time down to about an hour I think:

- Although I’m sure I don’t retain the words as well, I’ve stopped marking recent words “Hard”; after I initially study them and get them into short-term memory, I choose “Good” in Anki which means I see them 3-5 days later instead of 1-2. In fact overall I avoid marking anything “Hard”; it’s either “Again”, “Good”, or “Easy”. This has seemed to help cut down on the amount of daily reviews.

- I used to go over new words several times over in Anki cram mode – 5-10 times for each word or more – which took forever. Now as soon as I get a word into short-term memory, I stop drilling it in Anki. I do make a list of new words for a given day, post them on my computer at work and manually drill them throughout the day. But somehow that is less painful than being stuck in Anki drilling them.

- Previously I was only using Anki on my desktop. I now sync my deck with AnkiOnline and do my reviews via AnkiOnline using my iPod Touch. For whatever reason reviews are easier to stomach this way as opposed to having to sit at the computer and do them. I still learn new words on the desktop, which is fine.

- Although I am not studying sentences – I’m focusing on vocab – I was initially religiously shadowing the sentence audio for every word. Now I only do that while I’m learning new words. During reviews I skip the shadowing part to make things go more quickly.

- Initially I was using the cram feature to study all 20 new words at once. I now focus on a few words at a time, still drilling them using the cram feature. In other words I cram the 1st 5, then the 1st 10, then the 1st 15 and finally all 20. It’s a little less overwhelming that way as I only focus on 5 or so new cards at a time.

Those adjustments have made things a little better but 6000 words in a year is still a little too aggressive for me. I know myself, and I can’t see myself keeping up this pace. In order to finish out all 6000 this year I would have to be pretty consistent day to day. I have a buffer of about 50 days where I can skip learning new cards and still finish this year, which seems okay, but I’ve already found that I’m less likely to study over the weekend, meaning I’m missing about 2 days per week, if not 3. I got sick and burnt out and stopped studying for about a week recently. Enough of those kinds of instances and I’m not sure 6000 is realistic for me.

I considered skipping the kana to kanji production part and just doing recognition (kanji to kana), but… to be able to recognize a word but not write it from memory is not ideal to me. Writing has been such an important part of learning languages for me in the past, and to not have that component would be problematic for me (in principle; I know it has been argued that from a practical standpoint you don’t really need to know how to write Japanese by hand). But there’s no doubt that doing kana to kanji adds a considerable amount of time and mental effort.

All that said, I have decided to cut my goal back from finishing all 6000 words this year to getting through the cards tagged JLPT 2 this year; that would bring me to a total of 3730 words. If I can do more than that, great, but this way I hope I won’t feel so anxious about having to learn new 20 words religiously every single day. I can get sick, get lazy, go on vacation, etc., and still reach my goal. I still plan to shift my focus to native materials next year, so I would probably drop the remaining 2270 words (or maybe activate them as I came across them while reading? I don’t know. I’m not a big fan of Anki reviews and do not want to have to do them indefinitely). 3730 words won’t give me quite as much of a boost as 6000 would have, but… I’m on the brink of getting burnt out, so I have to adjust somehow.

@mgbp7

I'm 4500 words into core.
After the 3000 word mark it was getting gradually easier for me.


I calculated that if you have 5k reps to do, and it requires about 15 reviews per word to really remember it, then if you do 5 reps per minute...( My way of doing is here http://forum.koohii.com/showthread.php?tid=7097)

...then you will reach your goal of total 6k words if you do reps for 1 hour for 250 days.
It always takes longer than anki shows, maybe 1,5 or 2 hours( loss of time, distractions).

The good thing is that the reps are always getting gradually easier, but it's harder to predict because of this.

When my motivation was high, I have even done something like 30 reps per 2 minutes (my record), although right now I do an average of 15 reps per 2 minutes.
----------

But anyway, if you'd study 10 hours on Saturday, and 10 hours on Sunday (which is possible if you have motivation).
It would be 20 hours per 7 days.

Calculate the average loss of time, and it would be about 10-15 hours per 7 days!(If you study on sunday and saturday etc.)

So this way you'd even be able to do even more than the remaining 5k words in 250 days!


...well it all depends on your motivation and the time you can spend on it.

EDIT
I actually really recommend you to read a lot of hiragana. If you're doing the same core deck that I am, the fast recognition of hiragana & words will really speed up doing reps too.

Thanks jettyke for the suggestions. Phew, 10 hours on Saturday and Sunday... I admire anyone who can keep that kind of pace, but I think I'd burn out within a week or two.

As for hiragana, my hiragana recognition is okay; what I've noticed improving is my katakana recognition. I think this might be a combination of the Core deck and all the katakana I read while playing FFXIII.

One more question -- you say you're 4500 words into Core (congrats, by the way). Have you noticed a marked increase in your reading ability? Would you say it's been worth the effort?

Thanks again!

I finished Core6K yesterday and I have two comments: Core6K gets really easy after a couple thousand of sentences (after 3 or 4K it's really not challenging). You develop a natural ability to guess a lot of compounds and what not, and it isn't as hard to remember the new sentences because there are less unfamiliar elements.

My second comment is that Core6K leaves out way too many kanji. I've ranted a little bit about this before, but covering ~1600 kanji is great and all but it doesn't cut it. You will never find a single page of text that isn't loaded with compounds and kanji you don't know. I'm currently contemplating what the most efficient way is to get the remaining joyo kanji completed so I don't feel plagued by ignorance whenever i'm reading. As a note I'm just speaking of readings here -- of course developing an actual _vocabulary_ well beyond 10K words will be necessary for full comfort in reading, but it's SO frustrating to come across kanji compounds I have no idea how to read (referring to reading physically printed material here, where rikaichan can't save the day). Just my two cents.

k.

mgbp7 Wrote:Thanks jettyke for the suggestions. Phew, 10 hours on Saturday and Sunday... I admire anyone who can keep that kind of pace, but I think I'd burn out within a week or two.

It may be hard for you to imagine it right now, but at least for me it has changed a lot since I started (I didn't do 2k, because I knew all the basic words).

BTW I haven't done any other decks since starting this one for your information.

As you can see, the amount of reps done per day has changed a lot since starting.
As you notice, there are record high points and then drastic falls for 3 times. It means that I haven't been at a steady pace, but I think it's okay. So don't worry about doing them steadily every day!!!

And to be frank, since starting this deck my methods and motivation levels have changed greatly! So it might be a reason for the change in the amount of reps I've done per day. Don't worry about the numbers though, they're just numbers



All of these 7 motivating factors have increased greatly in the past months and that's why I've spent more time on Japanese. Motivation is everything.

1) I don't want to live with my family
2) I want to be more independent
3) As almost everything I enjoy and value is associated with Japan one way or another, I'm suffering a lot because of not living in Japan.
4) Getting closer to my dreams lies in Japan
5) I'm getting close to high school graduation, so now is a very important time in which the decisions I make will change my life by a great deal.
6) I've probably fallen in love with a Japanese person and the lack of my Japanese abilities and the fear of getting separated soon give me motivation. And I like Japanese women in general, I can tell from my experience.
7) I've had a deep interest/connection with Asia that has changed my way of living for 5 years now and if there's anything I know, it's that I definitely have to live there.

...my goal is to live in japan


Imo the best way to do reps is when your motivation is at your peak! You can do a lot more when your motivation is high...

mgbp7 Wrote:One more question -- you say you're 4500 words into Core (congrats, by the way). Have you noticed a marked increase in your reading ability? Would you say it's been worth the effort?

When I started core I already knew more than 2k words.

What has increased not only my reading ability, but writing also:

1) reading all 8 chobits mangas
2) exchanging lots of emails with my real life Japanese friends on mixi (something like 90 emails). It's was especially motivating that I exchanged the messages with people I really had an interest in, not just that "oh, they're japanese !!!".
3) Getting those emails corrected on lang8 before sending them.(it also reinforced core words)

Core

A lot of the words which I learnt from core popped up much more frequently than any of the other small decks or portions of words that I have ever encountered!
So in this sense I truly believe that core is the way to go.

You will never truly learn the words just by using an srs, never.

All that matters is what happens to the words you learn through SRS afterwards! All the REAL learning and improvement will take place after SRS!

... And as core has words which I encounter and use a lot, then I really think that core is a very good thing to srs. There will be a lot of recognition of at least having seen SOME certain words before!

Feel free to ask more, if you have any questions...

kodorakun Wrote:My second comment is that Core6K leaves out way too many kanji. I've ranted a little bit about this before, but covering ~1600 kanji is great and all but it doesn't cut it.

k.

I don't think it leaves out too many kanji. For example N2 tests for 6k words also, but it covers only 1000 kanji!

You can't expect such a small deck to contain a lot of kanji.

The only problem is that it's hard to find those remaining kanji, but it's not really a fault of core6k.

Thanks so much for your feedback kodorakun and jettyke. Congrats kodorakun on finishing up the Core deck.

Cool... even at 1200 or so I've noticed some of things you both have mentioned -- picking up on readings in a very natural way and seeing a lot of the Core words in native material.

But I also get what you're both saying about having realistic expectations; even if I finish the deck, I will still have a lot of work to do as far as vocab/kanji is concerned. Core can give a nice boost, a kind of introduction to common vocab and kanji compounds, but there will still be much to do afterward.

I accept the fact that it will be years until I'm able to read with any decent ease (and that only if I work at it) . As it is now I can't really read much of anything so much as "decode" it... look up vocab and grammar to figure it out.

To that end I've decided to try working through some native material (bilingual Doraemon comics to start with) and unsuspending unknown words in my Core deck as I go along (instead of just randomly going down the Core list with no context). This is my second day doing it this way and it's almost more work; I'm creating cards for words that aren't in the Core deck... and because Doraemon is kind of light on kanji usage (a sort of blessing but mostly a curse), if the word isn't already in Core I often have to look up the kanji information online and incorporate that manually.

We'll see how it goes. But like you said jettyke, motivation is important, and I'm hoping doing it this way will keep me more motivated. This way also breaks things up a bit... I plan to review every day but I'm not planning on adding/learning new words every day. I'll cycle through reading/decoding Doraemon, unsuspending/creating cards in Core, and actually studying the new cards in Anki. I have to pace myself, especially since I'm being dumb and studying Korean concurrently (though at a slower pace).

And finally, thanks for sharing your motivation jettyke... I'm really impressed with how determined and focused a lot of people on this forum are at such a young age. When I was in high school I did have (well, and still have) a certain love for French and did some things outside of class because I loved it so much but that was about it. Good luck on achieving your goals!

One last question for you both: how are you studying Core? (Make that past tense for kodorakun). Vocab, sentences? Kanji to kana? Kana to kanji? Also, what is/was your experience as far as failure rates? I find myself failing a good number of cards repeatedly on an ongoing basis. Some of the vocab sticks and some doesn't.

Thanks again.

jettyke Wrote:You will never truly learn the words just by using an srs, never.

All that matters is what happens to the words you learn through SRS afterwards! All the REAL learning and improvement will take place after SRS!

This has been discussed before on here, and I definitely agree.

Pre-made sentence decks are a crutch; a quick and easy to learn kanji readings and grammar patterns. However, you don't really *know* a word (i.e. you won't internalise it as word -> concept) until you have encountered it outside of your SRS, and recognised it *in context*.

I studied core6k by doing full kanji "mode" or whatever it was. Initially I had the sound prompt and would read after, but I found that too easy and didn't think I was really reading. So I switched my cards up to have the sentence alone as the prompt and I'd try to read the whole thing. The answer side was the audio (to shadow) and the english translation, along with any notes I'd add for the particular card (rarely ever added notes). I found that I was memorizing the entire sentences at one point, so I added a core6K deck that was a prompt with the single vocab word and the answer side having the audio and the english definition. It sounds like a lot of work but my vocab-only deck was extremely fast and really good at a) forcing me to really be able to define a word b) being able to read a compound word out of context c) see the vocab word without conjugation.

Maybe it was a bit overkill but my vocab-only deck enver took more than about 12 minutes on any given day to do the reviews -- it was really fast and on the answer side I also displayed the sentence (but no audio for the sentence) so I could check the contextual meaning.

I also did a small bit of resorting of cards. When I was at about 1200 unique kanji studied I suspended all cards and then systematically unsuspended 10 new cards per day that contained only unseen (i.e. new kanji). This way I finished all the ~1600 unique kanji kind of quickly and the remaining thousands of cards contained nothing but kanji I had _some_ exposure to, which made learning them quite easy. Be careful though, I think one could easily overdo it and take the extreme (i.e. first 1600 cards of Core6K study are the unique kanji cards would probably create more confusion than benefit). A good mixture of ~5 new kanji characters and ~10 cards that contain new vocab with already-seen kanji might be a stable base, adjustable according to your time and efforts.

A final note: My friend who stayed with the audio-prompt all the way through has developed a pretty good ear, so that may be a factor to consider. I don't know if my reading ability is better or worse, but a big part of core6k is the professional audio, so you may want to use that feature as a focus -- you can always generate audio-less sentences later on with relative ease.

k.

(please comment briefly on my stats for the core2k and the other.. misc deck )
Thanks in advance for any suggestions/corrections and advises! Here you go http://bit.ly/dJk4N4

I'm missing the Images for Core2k/6k in the kore list. Is there an extra thread for the images?