Hi
output of GIZA++ generates a file with extension ".A3.final.gz". The format
of the output for a sentence pair is given below:
# Sentence pair (1) source length 4 target length 3 alignment score :
0.00643931
wiederaufnahme der sitzungsperiode
NULL ({ }) resumption ({ 1 }) of ({ }) the ({ 2 }) session ({ 3 })
I want to know what is this alignment score. As far as I know GIZA++ do
word alignment and align source words to target words with some probability
i.e. translation probability. other things for which IBM Models talk about
are fertility, relative distortion and absolute distortion but all of
these things are for word to word not for a complete sentence pair.
So is this the sum or multiplication of all these probabilities for all
word correspondences of a sentence pair. how this single number gives
information about word alignment of a complete sentence pair. I really want
to know. please tell me. please please please
Thanks in advance
Jyoti Srivastava
On Sun, Aug 14, 2016 at 9:30 PM, <[email protected]> wrote:
> Send Moses-support mailing list submissions to
> [email protected]
>
> To subscribe or unsubscribe via the World Wide Web, visit
> http://mailman.mit.edu/mailman/listinfo/moses-support
> or, via email, send a message with subject or body 'help' to
> [email protected]
>
> You can reach the person managing the list at
> [email protected]
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Moses-support digest..."
>
>
> Today's Topics:
>
> 1. Link translated (Adel Khalifa)
> 2. Re: Link translated (Dingyuan Wang)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 14 Aug 2016 17:26:18 +0300
> From: Adel Khalifa <[email protected]>
> Subject: [Moses-support] Link translated
> To: [email protected]
> Message-ID:
> <CAL3Vmkc8pDV3a_QZAm9xDbAwrdrGqSMg6m6UM_
> [email protected]>
> Content-Type: text/plain; charset="utf-8"
>
> Hello All,
>
> How to prevent moses engine from translating links like "
> http://mailman.mit.edu/mailman/listinfo/moses-support"; or continuous
> strings like "you_are_using_moses_engine".
>
> Regards
> Adel
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: http://mailman.mit.edu/mailman/private/moses-support/
> attachments/20160814/2c240982/attachment-0001.html
>
> ------------------------------
>
> Message: 2
> Date: Sun, 14 Aug 2016 23:28:08 +0800
> From: Dingyuan Wang <[email protected]>
> Subject: Re: [Moses-support] Link translated
> To: Adel Khalifa <[email protected]>, [email protected]
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset=UTF-8
>
> Hi,
> You can write some regex to pre-process these strings, and transform
> them into
>
> <np translation="you_are_using_moses_engine">
> you_are_using_moses_engine</np>
>
> and set `-xml-input exclusive`
>
> ? 2016?08?14? 22:26, Adel Khalifa ??:
> > Hello All,
> >
> > How to prevent moses engine from translating links like
> > "http://mailman.mit.edu/mailman/listinfo/moses-support
> > <http://mailman.mit.edu/mailman/listinfo/moses-support>" or continuous
> > strings like "you_are_using_moses_engine".
> >
> > Regards
> > Adel
> >
> >
> > _______________________________________________
> > Moses-support mailing list
> > [email protected]
> > http://mailman.mit.edu/mailman/listinfo/moses-support
> >
>
> --
> Dingyuan Wang (gumblex)
>
>
> ------------------------------
>
> _______________________________________________
> Moses-support mailing list
> [email protected]
> http://mailman.mit.edu/mailman/listinfo/moses-support
>
>
> End of Moses-support Digest, Vol 118, Issue 26
> **********************************************
>
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support
