nile

A discriminative syntax-based word alignment package for statistical machine translation

Please visit the Nile project page on GitHub

http://jasonriesa.github.io/nile/

Nile is a supervised, discriminative word alignment package that can make use of arbitrary and overlapping features. Our current supplied language-independent featureset enables accurate models of word alignment as tested on Arabic-English and Chinese-English language pairs. You can easily augment training with your own set of features specific to whatever language pair you are working with.

For details, see:

  • Feature-Rich Language-Independent Syntax-Based Alignment for Statistical Machine Translation. (J. Riesa, A. Irvine, and D. Marcu). 2011. In Proceedings of EMNLP, pp. 497-507.

  • Hierarchical Search for Word Alignment (J. Riesa and D. Marcu). 2010. In Proceedings of ACL, pp. 157-166.

  • Note: Code related to our NAACL 2012 paper involving derivation traversal and inspection will be available soon in the next release!

The latest documentation is available on the downloads page: http://code.google.com/p/nile/downloads/list

Related tool: Check out Picaro, a simple command-line word alignment visualization tool. http://jasonriesa.github.io/picaro

Project Information

The project was created on Mar 19, 2012.

Labels:
wordalignment alignment translation machinetranslation nlp mt syntax