Skip to content
/ cdec Public
forked from redpony/cdec

Decoder, aligner, and model optimizer for statistical machine translation and other structured prediction models based on (mostly) context-free formalisms

License

Notifications You must be signed in to change notification settings

kho/cdec

 
 

Repository files navigation

cdec is a research platform for machine translation and similar structured prediction problems.

Build Status

System requirements

  • A Linux or Mac OS X system
  • A C++ compiler implementing the C++-11 standard (NEW)
    • Unfortunately, many systems have compilers that predate C++-11 support.
    • You may need to build your own C++ compiler or upgrade your operating system's.
  • Boost C++ libraries (version 1.44 or later)
    • If you build your own boost, you must install it using bjam install.
    • Older versions of Boost may work, but problems have been reported with command line option parsing on some platforms with older versions.
  • GNU Flex

Building from a downloaded archive

If your system contains the required tools and libraries in the usual places, you should be able to build as simply as:

./configure
make -j4
./tests/run-system-tests.pl

Building from a git clone

In addition to the standard cdec third party software requirements, you will additionally need the following software to work with the cdec source code directly from git:

Instructions:

autoreconf -ifv
./configure
make -j4
./tests/run-system-tests.pl

Further information

For more information, refer to the cdec documentation

Citation

If you make use of cdec, please cite:

C. Dyer, A. Lopez, J. Ganitkevitch, J. Weese, F. Ture, P. Blunsom, H. Setiawan, V. Eidelman, and P. Resnik. cdec: A Decoder, Alignment, and Learning Framework for Finite-State and Context-Free Translation Models. In Proceedings of ACL, July, 2010. [bibtex] [pdf]

About

Decoder, aligner, and model optimizer for statistical machine translation and other structured prediction models based on (mostly) context-free formalisms

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 89.4%
  • Python 4.8%
  • C 3.4%
  • Perl 2.3%
  • Other 0.1%