We try to optimize parameters using MERT and use reordering models to improve these results.
Table9 shows the results of these experiments. As can be seen this table, proposed method was not so effective for BLEU score. However, it was effective for METEOR score. It means that proposed method was less unknown words.