Hi Lucia.

No, it does not.

There are several reasons I can think of for this fact.

First, a "heuristical" reason in training time: if you take a look at  
how Moses does the phrase extraction, you will realize it has no sense  
to consider them.

On the other hand, a more theoretical reason in decoding time:  
considering 0-fertility segments (n-word source-phrases aligned to  
null) boosts complexity, and the problem becomes unfeasible, as far as  
I remember.

Regards,

Germán

Quoting Lucia Specia <[EMAIL PROTECTED]>:

> Hi all,
>
> Does Moses extract phrase alignments to 'null'? I mean, alignments
> between a n-word phrase and 0 target words?
>
> I had a look at the phrase tables and it seems that this does not
> happen, but I'd liket o confirm whether that is true.
>
> Thanks a lot,
>
> Lucia
> _______________________________________________
> Moses-support mailing list
> [email protected]
> http://mailman.mit.edu/mailman/listinfo/moses-support
>



----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.


_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to