Page 1 of 1

Never Ends = NeBa - ENJiSu ?

Posted: Sun Nov 02, 2003 6:01 pm
by The_Shape
Is the katakana for "never ends" = "ne ba - e n ji su" (neba enjisu)?

Oh! Great is english, but it's written in kanji, right? Oogure Ito?

Or am i confusing things? :?:

Posted: Sun Nov 02, 2003 6:04 pm
by FuguTabetai
Oh Great is written in Kanji. Ou gure hito

Never ends? I don't know, but that sounds unlikely. It would more likely be neba- endo- (the would just leave off the "s".)

fugu

Posted: Tue Nov 04, 2003 12:46 am
by The_Shape
FuguTabetai wrote:Oh Great is written in Kanji. Ou gure hito

Never ends? I don't know, but that sounds unlikely. It would more likely be neba- endo- (the would just leave off the "s".)

fugu
Thanks!

By the way, is there a rule to this, or is it just random?

Posted: Tue Nov 04, 2003 9:12 am
by FuguTabetai
There are rules, but they do not cover all cases, and often you just have to go by what is more popular/used.

fugu

Posted: Tue Nov 04, 2003 1:17 pm
by FuguTabetai
I couldn't pass up this opportunity to pass on (probably) more information than you would like, but it is relevant to the area that I do research in.

You can get a nice paper on automatic transliteration (statistical) from English to Japanese from the <a href="http://www.amtaweb.org/summit/MTSummit/ ... eedings</a> of the most recent <a href="http://www.amtaweb.org/summit/">Machine Translation Summit IX</a>. The paper is <a href="http://www.amtaweb.org/summit/MTSummit/ ... literation considering context information based on the maximum entropy model</a> by I. Goto, N. Kato, N. Uratani and T. Ehara.

You might want to brush up on your statistics, but with a word accuracy of about 70%, you can clearly see that there are "rules" - or at least patterns that can be modelled fairly well by statistical processes.

cheers,

fugu

Posted: Sat Nov 08, 2003 12:15 am
by The_Shape
Cool. Thanks, Fugu!