You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@opennlp.apache.org by "Jeff Zemerick (Jira)" <ji...@apache.org> on 2022/06/07 11:58:00 UTC
[jira] [Reopened] (OPENNLP-1306) NameSample overlap exception not helpful
[ https://issues.apache.org/jira/browse/OPENNLP-1306?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jeff Zemerick reopened OPENNLP-1306:
------------------------------------
> NameSample overlap exception not helpful
> ----------------------------------------
>
> Key: OPENNLP-1306
> URL: https://issues.apache.org/jira/browse/OPENNLP-1306
> Project: OpenNLP
> Issue Type: Improvement
> Components: Name Finder
> Affects Versions: 1.9.2
> Reporter: Markus Jelsma
> Priority: Major
> Fix For: 1.9.5
>
> Attachments: OPENNLP-1306.patch
>
>
> I got this for some very large training file.
> {code:java}
> Computing event counts... Exception in thread "main" java.lang.RuntimeException: name spans [27..29) person and [27..27) person are overlapped in file: null
> at opennlp.tools.namefind.NameSample.<init>(NameSample.java:79)
> at opennlp.tools.namefind.NameSample.<init>(NameSample.java:97)
> at opennlp.tools.namefind.NameSample.<init>(NameSample.java:101)
> {code}
> With this exception it is impossible to track the error if you have a large training file.
>
> Exceptions about mismatching <START:> and <END> tags at least give a little bit of context. This patch adds the sentence parts to the exception, making it simple to grep the training file for the bad sentence.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)