Strict Standards: Declaration of SkinTJP::initPage() should be compatible with SkinTemplate::initPage(OutputPage $out) in /var/www/ on line 0 Google - WagaWiki


From WagaWiki

Revision as of 09:10, 13 January 2007 by Paul b (Talk | contribs)
(diff) ← Older revision | Current revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Google can be a great way to check your Japanese and to investigate questions about the Japanese language. For example if you can't decide whether to use kanji or kana do a google on both versions and see which has the most hits.

However there are many traps waiting for the unwary Google user.

  • Google language selection
If you search for something that includes kanji but no kana the results will include Chinese pages. What's worse is that Google treats these searches differently to normal Japanese searches so sometimes it won't return the Japanese pages that should have hits. Always specify Japanese language (Google advance search) or include kana in the search.
  • Google splits phrases
If you search for a string of Japanese text Google will often decide to split up your search string into 'words'. For example if you search for 電車勉強[1] you will get pages that match both 電車 and 勉強 but don't have 電車勉強 as a single phrase. Always use quote marks around phrases of more than one word.
  • Google indexes on words
A search for 車 will not return pages that only include 電車 because 車 is only part of the word 電車. Also splitting up verb conjugations and such will result in poor search results. 話します returns over 1 million results but 話しま returns only 158,000 (NOT including 話します).
  • allintext:
Recently Google has started returning very high spurious hit numbers on some Japanese text searches. If you add "allintext:" to the start of your search string (or set "Return results where my terms occur" to "in the text of the page" you will be OK.

[1] Why you would want to I can't imagine.

  • Why Google?
Good question. I find that has more reliable hit number estimates and doesn't have as many quirks to deal with. Note however that doesn't handle "exclusion" searches well.
Update Recently I have found some searchs to occasionally return spurious very high numbers. I guess the only answer is that if you need to be sure you should use both Google and Yahoo.

Keep the above points in mind and have fun learning with search engines.

Personal tools