Joern Kottmann created OPENNLP-603:
--------------------------------------
Summary: Add a token cluster feature generator
Key: OPENNLP-603
URL: https://issues.apache.org/jira/browse/OPENNLP-603
Project: OpenNLP
Issue Type: Improvement
Components: Name Finder
Reporter: Joern Kottmann
Assignee: Joern Kottmann
Priority: Minor
Fix For: 1.6.0
Add a feature generator which can lookup the token cluster from a dictionary.
Depending on the data set this can increase the performance up to a few percent
f1 measure.
See for example this paper:
www.aclweb.org/anthology/P10-1040.pdf
A powerful tool to produce token clusters is for example:
https://code.google.com/p/word2vec/
--
This message was sent by Atlassian JIRA
(v6.1#6144)
