You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@opennlp.apache.org by "Joern Kottmann (JIRA)" <ji...@apache.org> on 2015/09/17 14:12:04 UTC

[jira] [Created] (OPENNLP-819) Leipzig corpus reader should be able to train a language identification model

Joern Kottmann created OPENNLP-819:
--------------------------------------

             Summary: Leipzig corpus reader should be able to train a language identification model
                 Key: OPENNLP-819
                 URL: https://issues.apache.org/jira/browse/OPENNLP-819
             Project: OpenNLP
          Issue Type: Improvement
          Components: Doccat
            Reporter: Joern Kottmann
            Assignee: Joern Kottmann
            Priority: Minor
             Fix For: 1.6.1


In its current state the Leipzig corpus reader can only read from one language file. In order to create a model that can detect many languages all the input files must be converted and merged together.

It would be much easier to train a language identification model if the corpus reader could just read many sentences files form a directory.

This issue will change the Leipzig reader to read from all sentences file in a specified directory. The language category should be extracted from the file name itself.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)