Re: Our plan to improve Calc functionality related tophonetic text

Takashi Nakamoto wrote:
> On Tue, 21 Aug 2007 14:50:15 +0200
> Eike Rathke <erlsun.com> wrote:
>
>   
>> Hi Kohei,
>>
>> On Tuesday, 2007-08-21 08:35:41 -0400, Kohei
Yoshida wrote:
>>
>>     
>>> IIRC, if the text cell doesn't contain any
phonetic text, it will
>>> back-translate it into ruby using the IME if
the main text contains any
>>> Kanji characters, then return that ruby text to
replace the Kanji
>>> characters in the main text.
>>>       
>
> It's not true.
>
> If a text cell doesn't contain any phonetic text, it
just returns base
> text without back-translation at least on my
environment (Window XP,
> Excel 2003).
>
> Back-translation, which is sometimes called
"reconversion", comes up
> when you change continuous Kanji characters which
compose one word.
> It's difficult to explain the detail of this situation
in English but
> I've introduce the understandable example following.
>
> The word "research" can be translated
"kenkyuu" in Japanese. This
> Japanese word is expressed by two Kanji characters,
[U+7814, U+7A76]
> and the corresponding phonetic text is [U+3051, U+3093,
U+304D,
> U+3085, U+3046]. Think of the situation that a cell in
Excel contains
> this word with the corresponding phonetic text.
>
> Then, split the base text [U+7814, U+7A76] with a space
character to
> make it [U+7814, U+0020, U+7A76]. At the same time, the
phonetic text
> changes. It becomes [U+3051, U+3093, U+0020, U+304D,
U+308F, U+3080].
> The last two characters change from [U+3085, U+3046] to
[U+308F,
> U+3080], because back-translation returns [U+304D,
U+308F, u+3080]
> when [U+7A16] is used as one word.
>   

That means the 'KUN' and the 'ON'-meanings (kunyomi vs
onyomi)? [I don't 
know much about ruby; nevertheless I struggle to learn
Japanese. IF I 
only had more time, ...a 25 hour day or something like that,
and 
somebody with whom to exercise...  :- ).

>> So to produce identical formula results with a
document on different
>> systems it also depends on the availablity of an
IME that has
>> back-conversion and works identical and offers that
through some API?
>> Sounds like a bunch of mess to me.
>>     
>
> In addition, because back-translation, probably stands
on difficult
> stochastic algorithm, IME returns different results in
rare cases. The
> result of back-translation also depends on environment
and
> configuration.
>
> It might sound like a bunch of mess to you. Yes, we
should realize
> that we are facing a difficult problem.

------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribesc.openoffice.org
For additional commands, e-mail: dev-helpsc.openoffice.org

> > Back-translation, which is sometimes called
"reconversion", comes up
> > when you change continuous Kanji characters which
compose one word.
> > It's difficult to explain the detail of this
situation in English but
> > I've introduce the understandable example
following.
> >
> > The word "research" can be translated
"kenkyuu" in Japanese. This
> > Japanese word is expressed by two Kanji
characters, [U+7814, U+7A76]
> > and the corresponding phonetic text is [U+3051,
U+3093, U+304D,
> > U+3085, U+3046]. Think of the situation that a
cell in Excel contains
> > this word with the corresponding phonetic text.
> >
> > Then, split the base text [U+7814, U+7A76] with a
space character to
> > make it [U+7814, U+0020, U+7A76]. At the same
time, the phonetic text
> > changes. It becomes [U+3051, U+3093, U+0020,
U+304D, U+308F, U+3080].
> > The last two characters change from [U+3085,
U+3046] to [U+308F,
> > U+3080], because back-translation returns [U+304D,
U+308F, u+3080]
> > when [U+7A16] is used as one word.
> >   
> 
> That means the 'KUN' and the 'ON'-meanings (kunyomi vs
onyomi)? [I don't 
> know much about ruby; nevertheless I struggle to learn
Japanese. IF I 
> only had more time, ...a 25 hour day or something like
that, and 
> somebody with whom to exercise...  :- ).

It's an off topic but I would like to tell you the reason. I
want as
many people to know about Japanese as possible.

Yes, the former is "ONYOMI" and the latter is
"KUNYOMI". Essentially,
this problem is caused by the fact that one Kanji character
has
generally multiple readings (pronunciation). Surely, the
readings can
be classified roughly into "KUNYOMI" and
"ONYOMI". However, don't rush
for a conclusion that every Kanji character has two
readings.

In general, one Kanji character has one reading, two
readings or more
than two readings. For example, 
http://www.kanji-a-day.com/dictionary/kanji-detail.
php?id=292
This is a Kanji character, which usually means eye. It has
four
readings following;
 - Gan
 - Gen
 - Manako
 - Me
The first and second readings are "ONYOMI" and the
others are
"KUNYOMI".

The choice of reading completely depends on the context. In
addition,
in rare cases, the reading can't be determined uniquely,
hence,
back-translation is based on a probabilistic algorithm. This
is why
back-translation seems a bunch of mess to you.

--
 Japanese Native-Language project
 Good-Day Inc.
 Takashi Nakamoto

------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribesc.openoffice.org
For additional commands, e-mail: dev-helpsc.openoffice.org