I wonder what would be the implications if a RDBMS would be used for the decoder. I;m also wondering what Google Translate uses – is there any public info about them in this respect?

Catalin Braescu
Omlulu.com


Re:[Moses-support] RDBMS for the decoder?

Hi,

I have no idea how easy it is to use conventional database to store the models.

Regarding Google translate: they use very similar methods, but it is implemented in a parallel way. If you are interested in that – check out Hadoop, or this tutorial:

http://clear.colorado.edu/NAACLHLT2009/tutorial_docs/DistLM.description.pdf

-phi

Re:[Moses-support] RDBMS for the decoder?

The decoder in the mt3_chart uses Berkeley Db to store the translation table. This it more to enable us to have a flexible platform to work with. Performance is ok, and can be tweaked if you know more about BerkeleyDb.

The decoder architecture allows you to easily add new translation table implementations.

I haven’ttried using RDBMS but it may be unacceptablilty slow as there are more layers, eg. compiling SQL, query optimisation, returning records.

NOTICE:This is digested from the Moses-support mailing list, which supports for the moses SMT decoder.

Related posts:

1. Moses Support Digest: Moses on the iPhone
2. Moses Support Digest:compiling moses 3 chart
3. Moses Support Digest:A translation chain prototype with Moses + IRSTLM
4. Moses Support Digest: Moses decoder on windows freezes after 5 sentences
5. Moses Support Digest:POS translation
6. Moses Support Digest:CreateBerkeleyPt and On-Disk Rule Table
7. Moses Support Digest: moses_chart and recaser phrase-table
8. Moses Support Digest:moses decoder results on cygwin and dos
9. Moses Support Digest:nbest list option in decoder
10. Moses Support Digest: moses install on 64-bit