Re: [tlug] a japanese dictionary

["Michael Engel" <mkengel@example.com>]

David.

I have no idea how you are going to implement Kanji searches. There is
a lot of research going on in this direction. How to search an unknown
Kanji efficiently.

E.g., have a look at the following page:
http://cantor.ish.ci.i.u-tokyo.ac.jp/~julian/kansuke2.html
Just play around with it - it's fun. They might licence it to you.

Another approach is undertaken bei Hans-Joerg Bibiko.
http://bibiko.de/
He spoke about it on a recent workshop in Tokyo. By choosing graphemes
and a database behind it which knows all graphemes in every kanji
(ca.6000 Kanji), it is a very fast system for choosing a Kanji without
having it at hand or knowing it. You choose a graphem and immediately
you get information on the remaining possibilities, i.e. all the
Kanjis with that graphem. You choose a second one - and most probably
you already have the Kanji.

It is very nice software. Hopefully, it's coming out soon.

Best regards
Michael

Chinese-German-Chinese dictionary HanDeDict
http://chdw.de


On 4/2/06, David Stibbe <dstibbe@example.com> wrote:
> Michael Engel wrote:
> >>> I'm working on building a japanese dictionary
> >
> > Principally that's interesting, could you please explain your concept
> > of the dictionary in some words.
> > What structure will it have, where do you get the data for the
> > dictionary, etc...?
> >
>
> Currently I'm in the planning phase.
> I am afraid I should have phrased myself more clearly before, for you
> might have misunderstood the goal of my project.
>
> The concept is to create a graphical dictionary utility for several
> platforms (linux first, then windows and hopefully mac after that).
> I think you thought I was going to create a new dictionary as EDICT and
> KANJIDIC ?
>
> >> How about classifying the words by subject, searching for synonyms or
> >> related words and for antonyms (oposite meaning words)
> >
> > Are such data available or do they need to be written from scratch ?
>
> I do not know if this exists, but I wouldn't be surprised if it didn't
> for the antonyms; for the synonyms and 'related words' one could make a
> work around.
>
> >> A sort would be pretty usefull too. Sort by furigana, by stroke count, by
> >> some dictionary number. Grouping by kanji.
> > For such data, see the KanjiDic maintained by Jim Breen
>
> Even though this data is available, the current utilities do not often
> provide a whole lot of choices when it comes to sorting ...
>
>  > An interesting project would be to set up an "open (extendable) SQL
>  > database" and "open source scripts" for it.
>  > I guess every good dictionary has such a database behind it anyway.
>
> Even if though falls outside the scope of my original project, that
> wouldn't even be such a bad idea; this would be interesting to do :P
>
> Kind regards,
> David
>
> --
> The TLUG mailing lists and web site are temporarily provided by OnJapan.net
> http://www.OnJapan.net/
> The TLUG server is hosted by Open Source Development Lab Japan.
> http://www.osdl.jp/
>
> To unsubscribe from this mailing list,
> please see the instructions at http://www.tlug.jp/list.html
>