1

I am new to OpenNLP and have used it to get some parsing sentences.

I saw in Google TensorFlow youtube videos that the Penn Treebank is old and sort of outdated. They have made another parsing model named SyntaxNet available as Open Source.

My question is that is it possible to use SyntaxNet Models in Apache OpenNLP libraries.

SyntaxNet is a part of TensorFlow repo in git and is Python based. I am a Java Developer.

Thanks in Advance.

Mark Rotteveel
  • 100,966
  • 191
  • 140
  • 197
Fr_nkenstien
  • 1,923
  • 7
  • 33
  • 66

1 Answers1

0

OpenNLP would pick any tagging or syntactic/semantic model as long as you create the training data yourself. In this case, you would need to train POS tagger with their set of tags and also OpenNLP chunker and / or parser to implement SyntaxNet.

That said, my personal take on this is that SyntaxNet is a very opinionated piece and there is no reason to use it for instead of Penn Treebank. By doing so, you're locking yourself in a Google's solution.

taleodor
  • 1,849
  • 1
  • 13
  • 15