5

I generally dislike using flashcards to learn vocabulary (or kanji) because their meaning changes so much depending on context. Names are an exception, and also really hard for somebody who hasn't grown up knowing people who they can associate with the kanji.

I was looking at the list of resources we have here at meta, but there wasn't anything that could be used to compile a list of names by frequency, except perhaps the physical version of BCCWJ.

There is a website that has a Flash-based quiz for 516 most common family names, but I'm looking for data that I can use to create my own flashcards. The site claims that it collected the names from telephone books and unless the author actually manually counted the frequency, there must be a digital version of the telephone book somewhere.

The requirements for the data:

  • Kanji
  • Reading (preferably hiragana or katakana instead of romaji)
  • Frequency coefficient OR a raw list of kanji with readings
2
  • I'm sure you've seen it already, but Wikipedia actually fulfills all those requirements, but it only has the top 100 names.
    – Blavius
    Jun 7, 2017 at 21:16
  • @Blavius Thank you, looking at the source Wikipedia used at least confirms that this data is probably not freely available. I also guess that the readings were added afterwards by somebody with knowledge of the most common readings. Out of curiosity I went back in time on that Wikipedia page and at ~2010 this came up: web.archive.org/web/20080116052540/http://park14.wakwak.com/… There are no readings but I'll see if I can find the most common readings for them somehow.
    – siikamiika
    Jun 7, 2017 at 21:56

3 Answers 3

Reset to default
4

I was not able to find a list of names ordered by the frequency of the reading, but there were several sites that offered a list of names ordered by the frequency of the name, counting all readings. The site I linked in comments didn't allow even non-commercial use, but 名字由来net did as far as I understand.

The readings provided by kanji.reader.bz don't include the obscure ones that are in JMnedict, so scraping it was a good choice. I assume that in case of multiple readings listed for one name the most common one is first. Example:

田中: たなか、でんちゅう

I have always read that name 「たなか」.

The results are here:

https://github.com/siikamiika/japanese-family-names

EDIT: For anyone interested, here's what the flashcards look like at the moment. I will probably add some more features later and upload the deck to ankiweb or provide a download link at GitHub.

12
  • I'm glad I saw this post. I've been studying Japanese names recently but I have no access to frequency lists, and I don't know how to generate them based on the data here. Can you offer any advice on how to get an Anki deck started which contains names based on frequency? Thanks !
    – kandyman
    Sep 24, 2017 at 2:47
  • Is there a way I can contact you? I have several things I'd like to ask.
    – kandyman
    Sep 26, 2017 at 16:41
  • @kandyman I can't test it right now, but If I remember correctly, you have to run the command create_deck.py --output filename.txt with Python 3 and import the output to Anki.
    – siikamiika
    Sep 26, 2017 at 19:30
  • I'm not good with code, but I got around the problem by importing the csv file as data in Excel and selecting the right encoding. Then I exported to notepad in a format that is accepted. So I got there in the end.
    – kandyman
    Sep 26, 2017 at 21:29
  • @kandyman I added a download for the flashcard file to GitHub. You can import it to Anki from File → Import.
    – siikamiika
    Sep 27, 2017 at 9:24
  • Thanks! I've started looking for lists of first names, but it's hard to find anything of good quality. There are Anki decks with that contain first names mixed with family names, so it would be hard to separate them. And they don't have frequency data of course. But I will keep looking.
    – kandyman
    Sep 27, 2017 at 11:32
  • By the way, do you know from what type of corpus your list was derived?
    – kandyman
    Sep 28, 2017 at 13:10
  • @kandyman It claims that it's from 2017 statistics and phonebooks. I didn't find any names from the statistics so I guess it's processed from phonebook names and addresses combined with prefecture population statistics. I guess the phonebook refers to this.
    – siikamiika
    Sep 28, 2017 at 13:37
  • I've started studying your family name list and it's great! So helpful to start with familiar names. I've also started studying a list of first names, but it's painstaking because (a) they are not ordered by frequency, and (b) I have to check each name individually on yomikatawa.com/kanji and then manually input it. So slow!
    – kandyman
    Oct 8, 2017 at 12:03
  • List of Top 100 names for boys and girls, but it's in romaji:studentsoftheworld.info/penpals/stats.php3?Pays=JAP
    – kandyman
    Oct 10, 2017 at 11:00
  • Just an update: I've been studying the names you compiled. It's certainly very helpful but there are instances where the first name is not the most common. An example is 藤本. Without a doubt, by far the most common reading is ふじもと, but your list has とうほん as the first reading. I guess the reading variations are not fully accurate, eventhough the kanji frequency data are. Still, your list is a good resource and I am grateful to have it.
    – kandyman
    Oct 27, 2017 at 19:11
  • @kandyman Yep, some of them are not accurate. The readings are ripped from here. I guess it's still better than having all the readings in JMNEDICT like here: jisho.org/search/%E8%97%A4%E6%9C%AC%20%23name
    – siikamiika
    Oct 27, 2017 at 19:34
0

Great work !

Do you have any plans to make a frequency list for first names too?

2
  • If I find a list in some form somewhere, I can try to turn it to a list that is easier to process further (flashcards etc.). Also, I could be wrong but I think that first names are even more ambiguous than family names.
    – siikamiika
    Sep 26, 2017 at 19:26
  • Yes that would be amazing. It is so difficult to find good resources on the readings of names. And it is such an important aspect of life in Japan, in Japanese literature, etc. I think it's a very underrated part of Japanese studies. Something which native speakers know but most learners of Japanese never get round to studying.
    – kandyman
    Sep 26, 2017 at 21:30
0

I found a source for frequency-based first names, but it's not a free resource. I intend to buy it and make an Anki deck with it. I will give a review at a later time. http://store.tofugu.com/shop/japanese-names-by-frequency

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .