Hi,

you should not generate a reordering file, i.e. you should not
use the option "-reordering" when running the train-model.perl script
for hierachical/syntax models.

-phi

On Wed, Feb 24, 2010 at 2:53 PM, Jie Jiang <[email protected]> wrote:
> Hi Philipp:
>
> Yeah, I just noticed. Thank you very much!
> Could you please tell me why there is still another reordering-table file
> generated from moses-chart, looks like:
>
> [X] " [X] , two ||| [X] , [X] , ||| 0.333333 0.333333 0.333333 0.333333
> 0.333333 0.333333
> looks very similar to that in moses msd reordering table
> Best regards!
>
> Jie Jiang
> CNGL, School of Computing,
> Dublin City University,
> Glasnevin, Dublin 9.
> Tel: +353 (0)1 700 6724
>
>
>
>
> 2010/2/24 Philipp Koehn <[email protected]>
>>
>> Hi,
>>
>> the relationship between the Xs is encoded as word alignments.
>> In your example that is "1-0 3-2".
>>
>> The reorderdering model used in phrase-based models cannot
>> be used for hierarchical models.
>>
>> -phi
>>
>> On Wed, Feb 24, 2010 at 2:45 PM, Jie Jiang <[email protected]>
>> wrote:
>> > Hi Philipp:
>> >
>> > Thanks a lot!
>> >
>> > So what I'm wondering is that: in Chiang's paper, his rules contains X1
>> > X2
>> > on both source and target side to represent potential reordering,
>> > however,
>> > in moses-chart, by looking at the phrase-table extracted:
>> >
>> > [X] [X] ||| " [X] " [X] " ||| [X] “ [X] ” 使 ||| 1-0 3-2 ||| 0.298507
>> > 0.00219001 0.026435 0.000359774 2.718 ||| 0.124074 1.40106
>> >
>> > there is no X1 X2 stuffs here, does that mean any combination of reorder
>> > of
>> > the non-terminals are supposed to be accepted in this rules?
>> > And additional reordering scores could be calculated from another
>> > reordering-table that is generated in the training phase?
>> >
>> > Best regards!
>> >
>> > Jie Jiang
>> > CNGL, School of Computing,
>> > Dublin City University,
>> > Glasnevin, Dublin 9.
>> > Tel: +353 (0)1 700 6724
>> >
>> >
>> >
>> >
>> > 2010/2/24 Philipp Koehn <[email protected]>
>> >>
>> >> Hi,
>> >>
>> >> the best reference would be David Chiang's Computational Linguistics
>> >> article on hierarchical phrase-based models.
>> >>
>> >> -phi
>> >>
>> >> On Wed, Feb 24, 2010 at 11:21 AM, Jie Jiang <[email protected]>
>> >> wrote:
>> >> > Dear all:
>> >> >
>> >> > Could you tell me where can I find the details of moses-chart
>> >> > reordering?
>> >> > It seems that the reordering table is quite different from that of
>> >> > moses.
>> >> > Is there any papers or documents for it?
>> >> > Thank you very much!
>> >> >
>> >> > Best regards!
>> >> >
>> >> > Jie Jiang
>> >> > CNGL, School of Computing,
>> >> > Dublin City University,
>> >> > Glasnevin, Dublin 9.
>> >> > Tel: +353 (0)1 700 6724
>> >> >
>> >> >
>> >> >
>> >> > _______________________________________________
>> >> > Moses-support mailing list
>> >> > [email protected]
>> >> > http://mailman.mit.edu/mailman/listinfo/moses-support
>> >> >
>> >> >
>> >
>> >
>
>

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

Reply via email to