Hi,
you are on the right track.
> Suppose the following entry in an n-best list:
>
> 4 ||| así . ||| d: 0 -0.619042 0 0 0 0 0 lm: -11.4288 tm: -4.84733
> -6.39323 -6.90676 -7.23185 0.999896 w: -2 ||| -2.40665
>
> * "4"
> -> the number of the sentence
> * "así ."
> -> the output sentence
> * "d: 0 -0.619042 0 0 0 0 0"
the first number is the distance-based cost (total number of word movements),
the next six are the lexicalized reordering model costs:
- backward: monotone, swap, discontinuous
- forward: monotone, swap, discontinuous
so here the only non-zero value is the lexicalized reordering log-probablitity
of the first phrase translated monotone regarding to the sentence start.
> -> No idea ¿?
> * "lm: -11.4288"
> -> I suppose "lm:" stands for language model (log probability).
> Is this correct?
yes.
> * "tm: -4.84733 -6.39323 -6.90676 -7.23185 0.999896"
> -> I suppose "tm:" stands for translation model. But,
> to which translation model correspond each different value?
Currently, five different phrase translation scores are computed:
* phrase translation probability φ(f|e)
* lexical weighting lex(f|e)
* phrase translation probability φ(e|f)
* lexical weighting lex(e|f)
* phrase penalty
http://www.statmt.org/moses/?n=FactoredTraining.ScorePhrases
> * "w: -2"
> -> I suppose "w:" stands from word penalty.
yes, there are two words in the output.
> * "-2.40665"
> -> weighted overall socore.
yes.
-phi
_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support
