Hi experts,
I have a question on lexical weighting from the paper: philipp koehn, och, marcu. Statistical Phrase Based Translation. On page 5, subsection 4.4 Lexical weighting, an example is given how to compute lexical weights. Figure three gives this example where I assume ## are word alignments: F1 f2 f3 Null -- -- ## E1 ## -- -- E2 -- ## -- E3 -- ## -- I am under the impression that f is always the source language and e the target language. Given the null word in the example I think this is right. But then, how can source word f2 be mapped to two target words? Viterbi alignments only allow each source word to be mapped to one target word. What's going on here? Or maybe the ## mean that if a probability is present in the lex table, we use it? If so, why do we need the alignments then? Thanks in advance for explaining. Sanne
_______________________________________________ Moses-support mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/moses-support
