Yesterday I learned my 3000th word in Core6k.

I'm not sure what to do from now on; I have read some native material in the past couple of weeks, and more often than not I can get the 'gist' of what is being said. This alone is a great success to me, and a catalyst to keep reading more.

Obviously there are tons of new words in every other sentence. I don't particularly mind this since I use Rikaisama but what I do mind is that I can't put them into Anki without hindering my Core6k progress.

I will keep reviewing the cards I currently have in the Core deck; the choice I have to make, however, is whether to keep learning new words from Core or dive into native material and import words I find interesting/important into Anki.

Now that I've written all of the above, the choice would seem obvious- shift towards native materials. That said, it seems that importing vocab to Anki has become harder after moving on to v2.0 since Rikaisama doesn't support it yet. I was thinking of making a text file where I copy/paste the sentences I come across and then import them to Anki using Yomichan. Suggestions are welcome though....

Also, can I move my "in review" Core words to this new deck without the due date info being lost in the process? I think I'll keep the deck around to cross-check whether words I come across are already there in which case I can simply move them to the other deck.

I would dive into native materials while reviewing what you already know, and yes you can transfer cards from deck to deck without scheduling information being lost.

Are you interested in visual novels? The combination of ITH + TA makes adding new cards extremely easy.

@Betelgeuzah

I had a look at the new Anki version on a seperate computer, converting my Anki decks and see how it feels like and personally I feel it just has changed too much for me. I did not know about the Rikai-sama issue you write about, but this makes it another reason more for me not to meddle around with Anki v2 -- just lost time for me at least, even though it has some nice new features, I must admit...

The following might not be much of interest to you or even worthless since you are using the new version, but you might be interested in how I did it, though

What I did was, going through core6k straight and adding cards to a seperate deck I collected during lurking around the Internets with Rikai-sama. Since core6k is based on frequency of words appearing in Japanese - correct me if I'm mistaken - I decided to consequently work on core6k and concentrating more on adding words in this deck. After finishing it (after 9 months or so) I concentrated on lurking and collecting more words. I hadn't seriously read any novel whatsoever at this point of time, I just started recently as I grew tired (finally, one might say haha) with living in the Anki-box. I'm a very patient learner, so it took me awhile

What I'm trying to say is basically, since core6k was created based on frequency of words, stick to that first and don't worry too much about the rest. Try concentrating on this deck, see how it goes. Perhaps you will recognise changes (i.e. you understand more of what you hear when you listen to Japanese etc.), perhaps not. Then you can change the method again and see if there is something else suiting you better in terms of learning.

Thank you both. Sounds like converting my cards in review to a new deck should not be a problem.

Surprisingly (or not), even though I don't read native materials a lot yet (maybe a news article or two, or few pages of forum discussion in JP and so on each day) I constantly see words that I have learned in core or would encounter there very soon.

This makes me want to a) learn more words (in core) but also b) read more native material. However, not being able to "study" the new vocab I come across is discouraging. On the other hand, core really introduces high-frequency words so the progress I am making each day is very clear.

I feel like my grammar knowledge is at the stage where I don't see myself being flabbergasted by most sentences I come across. Thus I'm a bit tired of reviewing my DoBJG deck and it's sentences, since I "get" the grammar for the most part at this point. I really want to move on from pre-made sentences, but that won't happen unless I start reading native material.

Sadly I have come to appreciate the other features of Anki2 not to mention the sync feature of 1.0 is going away soon...it's not a real option for me anymore to use it.

What I was also thinking - is to learn 20 new words a day....consisting of however many words I come across and put into Anki + the remainder coming from the Core deck. If one day I will add only 5 words, the remaining 15 will be added from Core etc. Just an idea.

Either way, thanks for your opinions and insight.

Talking about grammar, I tried the DoBJG deck and it was ways to huge for me. I knew I would not make it and ended this short but insightful experiment.

I went through Tae Kim (reading the pdf countless times) and Genki 1 and 2; until now, nothing else in terms of grammar related studies has happened, however, I constantly looked up grammar terms I came across and could not handle in Kanzen Master, Donna toki and so on.

I have days I collect up to 170 new words. It entirely depends on how much I am in touch with native materials. E.g. I listened to 太田光 babbling in the background and caught the word 逆切れ, which I did not know and saved with Rikai-sama. Many more words I did not know where used by him in his speech, so... I started collecting. I could never add all the words at once to my Anki deck I encountered on one day lol Total overkill.

I'm currently doing Core 6k while reading native materials. I have definitely seen progress in my reading ability since starting it, but I don't really learn new words I see (in articles, books, games, etc), I just glance at the definition and move on. At my current new-vocab pace it'll take me about 70 more days to be done with Core 6k... after which, I was thinking about maybe making word frequency lists from things I watch / read & learning the more common non-Core ones. eh.. maybe I'll come up with something else in the coming 70 days. *still waffling through ideas*

Well anyway. As for Rikaisama, I'm pretty sure you can use it to save sentences into a Anki-ready import file. (Options -> Clipboard & Save -> Save Format -> something with $s )

The problem with a frequency deck is that the most important words are often not the most frequent ones. You could just move all the remaining core-cards to a "dump" deck and at the end of the day, add core cards if you haven't gathered much on your own, or to fill up your "daily word quota". This way, you can still gather words from native content and also learn more frequent core words. You could also add the core sentence with the word if the native sentence is too overwhelming. There are loads of ways of improving your vocabulary; just try different methods and see what works best for you.

And you could also just make a notepad-file with sentences and definitions ready to be copied into anki one by one. This is the least time-consuming way of "setnence mining".

Betelgeuzah Wrote:YNow that I've written all of the above, the choice would seem obvious- shift towards native materials.

It seems obvious because it is. Reading will help you progress in a way Anki can't: by providing a rich context, allowing you to understand the language far better than you ever could from just example sentences.

You should try reading a novel, without too much sentence mining. Or, instead of sentences, just take words that you commonly encounter in the specific novel you're reading, and add them to a new Anki deck (you could install Anki 1 in a different location on your hard drive, and use it just for this one purpose). Just make sure to keep the vocab deck light, with only the words that you absolutely need to follow the book you're reading: learning tons of vocab out of context is not the best use of your time and energy.

Besides, you can always just get back to adding to your Core deck, if you decide that reading novels is not for you. Another alternative you might want to look into is MCDs (as described on AJATT - including the free version, here: http://www.alljapaneseallthetime.com/blo...cds-part-1 -note part six, about the awesome Anki plugin that makes it easy to make the cards). These are of course best used in conjunction with reading a novel, and mining entire paragraphs out of it. [Note: these don't contradict the principle of keeping cards simple and focused on one piece of information, because the "massive context" is there just as a hint, it's not part of the question - you don't need to read it, you can just focus on the immediate context that contains the hidden character - therefor, even with that massive sentence, answering a card should not take longer than 10 seconds]

Thanks for all the replies.

I thought about this overnight and have come to the conclusion that I'll keep progressing in core while reading native materials as much as possible. I will also keep adding new words related to my interests to a notepad file, ready to be imported to Anki. Basically stuff that surely won't show up in my core deck but are high-frequency words anyway. Like that I hope to have quite a few words ready in time for when I "graduate" from core.

I'll be able to feel like I am "learning" new words while immersing myself (building the deck slowly but surely), not to mention attaining higher level of comprehension on the vocab and grammar I already "know", making my struggles more worthwhile in that regard.

[RawToast]

Stansfield123 Wrote:You should try reading a novel, without too much sentence mining. Or, instead of sentences, just take words that you commonly encounter in the specific novel you're reading, and add them to a new Anki deck (you could install Anki 1 in a different location on your hard drive, and use it just for this one purpose).

I did this with some graded readers (I am only at ~1100 words) and in Anki 2 I created a sub-deck 'Vocab:Books'. I found the words in a Core6k list and pulled them from my Core deck. After completing all the new words I just merged them into my Core deck. Those books no-longer trouble me with Vocab and briding those words 'forward' was worth it, as many of them were listed around positions 4000-6000.

Betelgeuzah Wrote:Now that I've written all of the above, the choice would seem obvious- shift towards native materials.

My #1 recommendation is subs2srs.

ryuudou Wrote:Are you interested in visual novels? The combination of ITH + TA makes adding new cards extremely easy.

My #2 recommendation would be vn2srs once released. Until then I hearily recommend ryuudou's suggestion. For anyone that's unaware, ITH extracts the text from the game and TA helps parse out words and lookup definitions.

RawToast Wrote:I did this with some graded readers (I am only at ~1100 words) and in Anki 2 I created a sub-deck 'Vocab:Books'. I found the words in a Core6k list and pulled them from my Core deck. After completing all the new words I just merged them into my Core deck. Those books no-longer trouble me with Vocab and briding those words 'forward' was worth it, as many of them were listed around positions 4000-6000. [emph. mine]

You can use the MorphMan plugin to automate this. Feed it a text file of the words you want to learn (could even just copy/paste the graded reader if you have an electronic copy) to create a morpheme database of words you want to learn. Then use the 'Mass Tagger' feature to tag all facts in your Core deck that contain one of the morphemes in the database. Then use Anki's study by tag feature.

Betelgeuzah Wrote:Also, can I move my "in review" Core words to this new deck without the due date info being lost in the process? I think I'll keep the deck around to cross-check whether words I come across are already there in which case I can simply move them to the other deck.

Not sure on the best way to do this. You can tag the words in this new deck that you already know from your Core deck (this time create the db by selecting all known Core cards and using Extract Morphemes feature), but I'm not sure how to retain due date info. Maybe you could "study" those tagged words and spam "easy" to sort of fake it. The real solution is to modify the Anki database if you're not afraid of doing a little SQL.

ryuudou Wrote:I would dive into native materials while reviewing what you already know, and yes you can transfer cards from deck to deck without scheduling information being lost.

Are you interested in visual novels? The combination of ITH + TA makes adding new cards extremely easy.

what is ITH+TA ??

errtu Wrote:what is ITH+TA ??

Two Windows programs: Interactive Text Hooker (for extracting text from games) and Translation Aggregator (for translating it). See http://www.anime-sharing.com/forum/tech-...ide-67016/.

overture2112 Wrote:My #1 recommendation is subs2srs.

do you use the video file to create the sound for it too? cause subs2srs seem to be super useful man, but to be looking all the time for vids that are in sync with the subs files sounds like too much trouble to me. So, the question is, how do you use it? cheers

edit. One more thing, rikaisama has been updated, it is now compatible with anki 2.0

errtu Wrote:do you use the video file to create the sound for it too? cause subs2srs seem to be super useful man, but to be looking all the time for vids that are in sync with the subs files sounds like to much trouble to me. So, the question is, how do you use it?

I don't personally use it but from what I understand, it's really easy to use with drama/subs from d-addicts and anime from kitsuneko.

For subs2srs, you have to mess about with the delay settings and pad-settings to make it right, but it is worth it. Usually, the same settings also works for other episodes of the same show.

errtu Wrote:

overture2112 Wrote:My #1 recommendation is subs2srs.

do you use the video file to create the sound for it too?...looking all the time for vids that are in sync with the subs files sounds like to much trouble to me.

I use video clips, audio clips, static image, and 2 lines of context before and after.

I edit all the subs files in Aegis to make my English and Japanese versions match up. You don't need to be perfect (subs2srs is fairly smart) but I usually have to fix up initial delay, delays for commercials (often the English fansubs assume no commercial and Japanese subs do), a song going on in the background, etc. As someone said, once you do one episode for a show, other episodes are probably similar, and once you do a few shows, you start getting pretty fast since you know what to look for and how to use Aegis.

It's important to remember that by fixing up the subs, not only do you get to benefit from subs2srs making extremely nice multimedia cards for you, but you also get to watch the video with Japanese subs (or better yet, use Morph Man's Adaptive Subtitles feature), dueling subs, etc.

I prefer to use subs2srs without English subs, which simplifies the process -- I only add Japanese definitions to the cards I end up using, which is about 10% of each "subs2srs dump".

Before I quit completely watching E-subbed material, I noticed that sometimes if a sentence was split into two sublines, the Japanese and English parts would not match, instead they would be like:

J -> E
A, B -> B, A

EDIT: ...and the video feature only crashed Windows Aero and didn't show any video at all...

Stian Wrote:...I noticed that sometimes if a sentence was split into two sublines, the Japanese and English parts would not match, instead they would be like:

J -> E
A, B -> B, A

EDIT: ...and the video feature only crashed Windows Aero and didn't show any video at all...

Ah, dealing with really long lines where one language's subs breaks them up and the other doesn't is another thing I do when fixing subs, which I think helps subs2srs avoid such mistakes.

I'm surprised you have an issue with playing video. subs2srs uses ffmpeg and Anki uses mplayer (which is based on ffmpeg).