Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: neutral (nike.apache.org: local policy)
Message-ID: <47F37465.70201@garambrogne.net>
Date: Wed, 02 Apr 2008 13:56:21 +0200
From: Mathieu Lecarme <mathieu@garambrogne.net>
User-Agent: Thunderbird 2.0.0.12 (X11/20080213)
MIME-Version: 1.0
To: java-user@lucene.apache.org
Subject: Re: stemming in Lucene
References: <329a6c830804010258jd8df420lcf0f777df9fc36f3@mail.gmail.com>
In-Reply-To: <329a6c830804010258jd8df420lcf0f777df9fc36f3@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit

Wojtek H a �crit :
> Hi all,
>
> Snowball stemmers are part of Lucene, but for few languages only. We
> have documents in various languages and so need stemmers for many
> languages (in particular polish). One of the ideas is to use ispell
> dictionaries. There are ispell dicts for many languages and so this
> solution is good for multilingual environment. Maybe this is not
> perfect place to ask, but does anyone know about java stemmer using
> ispell dicts?
> There is aspell-like java spell-checker (Jazzy) but I could not see
> how to use it for stemming. We are considering porting part of
> postgres tsearch module to java, because tsearch uses ispell dicts for
> stemming.
> But maybe there is a better way or there are people working on
> something like that?
>   
ispell data is nice for phonetic, and for enumerate a huge list of 
words. The ispell dictionnary is one way : pseudo root => word, it looks 
hard to build the inverse function, lemme is splitted in multiple affix. 
But it can be used to find rules, just like 
http://www.getopt.org/stempel/ do.

M.

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org