opennlp-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ian Jackson <Ian_Jack...@trilliumsoftware.com>
Subject English Training Corpus
Date Tue, 21 May 2013 15:06:56 GMT
As far as I can tell, only the results of the training are available for download. It appears
that the only method to change the model is to replace the model with a new model.

What corpus was used as the source for training the English models?
Was the Reuters Corpus used [http://trec.nist.gov/data/reuters/reuters.html]? If so in theory
an organization could sign the correct agreements and run the existing models to create something
close to input models and make the desired changes.


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message