lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Christian Moen (Commented) (JIRA)" <>
Subject [jira] [Commented] (LUCENE-3767) Explore streaming Viterbi search in Kuromoji
Date Tue, 28 Feb 2012 09:41:48 GMT


Christian Moen commented on LUCENE-3767:


Thanks a lot for this.  I've been taking this patch for a spin and I'm seeing some exceptions
being thrown when indexing some Wikipedia test documents.

I haven't had a chance to analyze this in detail, but when indexing the attached {{SolrXml-5498.xml}}
I'm getting:

java.lang.ArrayIndexOutOfBoundsException: -69
	at org.apache.lucene.util.RollingCharBuffer.get(
	at org.apache.lucene.analysis.kuromoji.KuromojiTokenizer.computePenalty(
	at org.apache.lucene.analysis.kuromoji.KuromojiTokenizer.computeSecondBestThreshold(
	at org.apache.lucene.analysis.kuromoji.KuromojiTokenizer.backtrace(
	at org.apache.lucene.analysis.kuromoji.KuromojiTokenizer.parse(
	at org.apache.lucene.analysis.kuromoji.KuromojiTokenizer.incrementToken(
	at org.apache.lucene.analysis.kuromoji.TestKuromojiTokenizer.doTestBocchan(
	at org.apache.lucene.analysis.kuromoji.TestKuromojiTokenizer.testBocchan(

I believe you can reproduce this by changing {{TestKuromojiTokenizer.doTestBocchan()}} to
read the attached {{SolrXml-5498.xml}} instead of {{bocchan.utf-8}}.

> Explore streaming Viterbi search in Kuromoji
> --------------------------------------------
>                 Key: LUCENE-3767
>                 URL:
>             Project: Lucene - Java
>          Issue Type: Improvement
>          Components: modules/analysis
>            Reporter: Michael McCandless
>            Assignee: Michael McCandless
>             Fix For: 3.6, 4.0
>         Attachments: LUCENE-3767.patch, LUCENE-3767.patch, LUCENE-3767.patch, LUCENE-3767.patch,
SolrXml-5498.xml, compound_diffs.txt
> I've been playing with the idea of changing the Kuromoji viterbi
> search to be 2 passes (intersect, backtrace) instead of 4 passes
> (break into sentences, intersect, score, backtrace)... this is very
> much a work in progress, so I'm just getting my current state up.
> It's got tons of nocommits, doesn't properly handle the user dict nor
> extended modes yet, etc.
> One thing I'm playing with is to add a double backtrace for the long
> compound tokens, ie, instead of penalizing these tokens so that
> shorter tokens are picked, leave the scores unchanged but on backtrace
> take that penalty and use it as a threshold for a 2nd best
> segmentation...

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message