|
BerkeleyAligner
Welcome to the BerkeleyAligner code repository. The version in the repository is 1.1, which fixes some issues with searching for posterior thresholds, and also changes the output format to match Moses. Note that the latest version is 2.0. Both source and pre-compiled versions are available from the Downloads tab. The source code for version 2.0 is currently not in the repository. |
► Sign in to add a comment
I want to use the aligner to compare with giza++ on speed and quality!
thanks.
Is there a documentation on how to use the .jar file in non-commandline native Java environment? http://stackoverflow.com/questions/8702724/how-to-use-berkeleyaligner-in-my-own-java-class
i am currently working on parallel corpus. can anyone suggest me a good tool to get word correspondences from sentence aligned corpora?
Hi, I am a traditional translator/translation scholar and have no experience using these machines - my area is Philosophy...
I have managed to make the unsupervised aligner work (I hope), but it does not produce this kind of file: union-soft-1+2.alignOutput.txt
It ouputs everything else, but the ones I so desperately yearn to see, don't appear.
Any idea what I am doing wrong?
Many thanks M.
Hello, I am working on moses, and I need to know whether to use giza or berkeley, that perform best alignment for languages with different syntax structures. that is between "subject verb object" and "subject object verb".
Thank you.