0

I'm trying to use statistical machine translations in python, and so far, I've been mostly using translate library. However, I heard that Moses, Phraser do a better job of machine translations, so I was wondering if there are ways of using them in python.

Is it possible?

Any explanation would be greatly appreciated.

user98235
  • 830
  • 1
  • 13
  • 31

1 Answers1

0

If you want to use in windows using Jupyter, you can use mosestokenizer https://pypi.org/project/mosestokenizer/ . But, it is a wrapper for Perl script and if it fails in Jupyter, use sacremoses https://pypi.org/project/sacremoses/.

However, if it is Linux/Unix, better to use tokenizer.py from https://pypi.org/project/mosestokenizer/#files with necessary arguments.

MrRaghav
  • 335
  • 3
  • 11
  • does it do translation? looks like it only does tokenization. – user98235 Jun 25 '20 at 20:10
  • I don't know if this satisfies your query but the following link says that a python interface for the decoder is still a project in Moses: http://www.statmt.org/moses/?n=Moses.GetInvolved . But, there are other blogs in which we can see python doing its work in translating the text, like - https://machinelearningmastery.com/develop-neural-machine-translation-system-keras/ – MrRaghav Jun 26 '20 at 06:41