![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
snarpI said I wasn't depressed about my JLPT score - but this was obviously all lies.
I'd been lazy about my kanji flash cards the last couple months, because I'd been using Mnemosyne and making my own hyper-specialized Remembering the Kanji cards. I was making my own cards, despite the existence of freely-downloadable pre-made decks for Anki, because I have special flashcard needs because I am an irrational, finicky person whose fingernails are always clean.
No one sane cares about my terrible slapdash process here, but I was, in brief, 1) learning a couple chapters, 2) compiling them into a deck of pure kanji flash cards, and then 3) making a set of vocabulary cards by going through Edict to find "Top 20,000" words containing only kanji already in my first deck. I was doing this all by hand, either in Mnemosyne itself or in Excel depending on my mood, so it was taking me nearly as long to add a new chapter as it was to go over it.
Please note that at least this time I did not go, "Well, the most efficient way for me to create the cards I need is to learn regular expressions first." Please note also that that last sentence contains the words "this time."
Also, Edict's "Top 20,000" list isn't all that great. So, on the whole, not efficient! And because, as previously stated, I am irrational and finicky, when I feel my study methods are inefficient I tend to lose motivation mid-way through a review and start writing a paranormal romance novel about orcs. I believe that the orc thing is common in these sorts of situations.
So yesterday I told my brain it would have to choose one half of the process to be anal-retentive about - 1) wasting huge amounts of time making my own flash cards, or 2) studying using cards that aren't perfectly suited to its preferred study methods. Both were not possible. I got the blue screen of death a couple times, but it eventually chose the first. So I downloaded Anki and one of the like, eight premade RtK kanji decks. That's working fine!
I don't want to talk about how many hours I spent making my own kanji cards before. Let's not discuss that.
Now the issue is adding in the vocab cards. Obviously I had to do research on how to do this most efficiently. I found a really interesting thread on the Koohii.com forums containing these statistics:
"173 kanji make up 50% all kanji in Wikipedia.
454 kanji cover 75% of all kanji in Wikipedia.
874 kanji cover 90%
1214 kanji cover 95%
2061 kanji cover 99%
2456 kanji cover 99.5%
3489 kanji cover 99.9%"
That right there? Useful information! Particularly given that shang, the person who compiled this data, cross-indexed the kanji in question with RtK and put them into a spreadsheet, which is on Google Docs here. (Also, one Katsuo added KANJIDIC indexes here, while one wonderflex color-coded the list by RtK section here.)
At one point there were word lists up to go with that, but the domain hosting them appears to have expired and been grabbed by squatters, and they're no longer up. These lists would basically be ideal for my vocab card purposes! I must find them.
(DO YOU SEE MY PROBLEMATIC HABITS RESURFACING HERE)
I'd been lazy about my kanji flash cards the last couple months, because I'd been using Mnemosyne and making my own hyper-specialized Remembering the Kanji cards. I was making my own cards, despite the existence of freely-downloadable pre-made decks for Anki, because I have special flashcard needs because I am an irrational, finicky person whose fingernails are always clean.
No one sane cares about my terrible slapdash process here, but I was, in brief, 1) learning a couple chapters, 2) compiling them into a deck of pure kanji flash cards, and then 3) making a set of vocabulary cards by going through Edict to find "Top 20,000" words containing only kanji already in my first deck. I was doing this all by hand, either in Mnemosyne itself or in Excel depending on my mood, so it was taking me nearly as long to add a new chapter as it was to go over it.
Please note that at least this time I did not go, "Well, the most efficient way for me to create the cards I need is to learn regular expressions first." Please note also that that last sentence contains the words "this time."
Also, Edict's "Top 20,000" list isn't all that great. So, on the whole, not efficient! And because, as previously stated, I am irrational and finicky, when I feel my study methods are inefficient I tend to lose motivation mid-way through a review and start writing a paranormal romance novel about orcs. I believe that the orc thing is common in these sorts of situations.
So yesterday I told my brain it would have to choose one half of the process to be anal-retentive about - 1) wasting huge amounts of time making my own flash cards, or 2) studying using cards that aren't perfectly suited to its preferred study methods. Both were not possible. I got the blue screen of death a couple times, but it eventually chose the first. So I downloaded Anki and one of the like, eight premade RtK kanji decks. That's working fine!
I don't want to talk about how many hours I spent making my own kanji cards before. Let's not discuss that.
Now the issue is adding in the vocab cards. Obviously I had to do research on how to do this most efficiently. I found a really interesting thread on the Koohii.com forums containing these statistics:
"173 kanji make up 50% all kanji in Wikipedia.
454 kanji cover 75% of all kanji in Wikipedia.
874 kanji cover 90%
1214 kanji cover 95%
2061 kanji cover 99%
2456 kanji cover 99.5%
3489 kanji cover 99.9%"
That right there? Useful information! Particularly given that shang, the person who compiled this data, cross-indexed the kanji in question with RtK and put them into a spreadsheet, which is on Google Docs here. (Also, one Katsuo added KANJIDIC indexes here, while one wonderflex color-coded the list by RtK section here.)
At one point there were word lists up to go with that, but the domain hosting them appears to have expired and been grabbed by squatters, and they're no longer up. These lists would basically be ideal for my vocab card purposes! I must find them.
(DO YOU SEE MY PROBLEMATIC HABITS RESURFACING HERE)
no subject
Date: 2010-03-14 04:44 am (UTC)*Ask me about ending charts sometimes! Besides the cases, there are male/female/neuter/plural and hard/soft.
no subject
Date: 2010-03-14 02:55 am (UTC)no subject
Date: 2010-03-14 03:57 am (UTC)(WOULD YOU ALSO LIKE MY THOUGHTS ON YAOI)
What are yours?
no subject
Date: 2010-03-14 04:38 pm (UTC)I think edict's kind of a dump. It's kind of a pre-wiki dump, managed by far fewer people (which, yes, it is smaller/more specialized), and less strictly moderated. I think one of the big problems is just what you said - definitions are usually (usually) technically accurate, but sometimes not made clearly enough dissimilar from one another. See: 'excited' vs. 'excited' (sexually), as learned through edict+j-go homework.
The recent formatting change makes it nice because everything is there, like a good dictionary, but not nice because it makes it much harder to copy/paste, like a bad online resource.
The short: edict is not quite accurate enough for me to use every day/without verification from a second source (which may be easily obtainable because it's online and you can google-search for whatever word you're looking at). It's kind of like wiki's less accurate cousin, and while I'll use it for reference, if I'm doing anything serious I'm probably going to pull out my heavy-duty paper dictionaries.
no subject
Date: 2010-03-17 07:54 pm (UTC)I will never trust Edict again.