7

Are there any equivalents of Google Ngram for Japanese?

Google Ngram doesn't currently cover Japanese, and the only other thing I could find was http://www.ldc.upenn.edu/Catalog/catalogEntry.jsp?catalogId=LDC2009T08 , which seems to be a paid-for service.

Golden Cuy
  • 16,209
  • 15
  • 74
  • 184
  • 1
    I don't have an answer, but what are you going to do with it? By the way, if you want to parse Japanese sentences, then ChaSen: http://chasen-legacy.sourceforge.jp/ is famous. –  Feb 04 '12 at 06:00
  • 1
    @sawa: I was wanting to find out if the word ソープランド existed before 1984. My reasoning was that if the word was the result of a "nationwide contest", then it wouldn't have existed beforehand, whereas if that was an urban legend, I'd see uses pre-1984. In the end, I used the English Ngram for the word soaplands. Conclusion: nationwide contest theory plausible. – Golden Cuy Feb 04 '12 at 06:23
  • Andrew: For that purpose, you can just do an ordinary search, you don't need N-gram. And that story is not an urban legend. It is a famous established story. –  Feb 04 '12 at 06:27
  • 2
    Yeah, unfortunately I don't know of an Ngram equivalent, but for this particular question there are many sources repeating the 1984 date, and a "Google Books" search finds nothing earlier: http://www.google.com/search?q=%E3%82%BD%E3%83%BC%E3%83%97%E3%83%A9%E3%83%B3%E3%83%89&hl=en&sa=X&ei=5-osT5TuLcL0mAXcpq3QDw&ved=0CBcQpwUoBA&source=lnt&tbs=cdr%3A1%2Ccd_min%3A1%2F1%2F1900%2Ccd_max%3A1%2F1%2F1986&tbm=bks – Matt Feb 04 '12 at 08:26
  • It seems like they've got it working for Chinese, so I wonder why they can't get it going for JP. I wonder if fuzzy word boundaries are a greater problem in JP than in ZH. –  Sep 24 '15 at 14:15
  • As a programmer and NLP specialist, I could easily write a similar program to google ngram. However, I would need an enormous corpus, the size of Google Books to work with. OCRing kanji is often very difficult, and I'm not sure there's much in the way of a Japanese Google Books equivalent to use. tl;dr: give me enough data and its totally do-able. – MatthewRNYC Feb 09 '12 at 16:40

1 Answers1

9

Google Japan sells the Japanese version of "Google Ngram". Here is the site: http://www.gsk.or.jp/catalog/GSK2007-C

But the site is only in Japanese. There might be the same site in English, though. I hope it helps.

senshin
  • 5,645
  • 30
  • 72
Hiromi Oyama
  • 106
  • 1
  • 2