Abstract:
In this study, we investigate English-to-Myanmar
statistical machine translation (SMT) by using
part-of-speech (POS) as linguistic inform-ation
which is valuable for statistical machine translation
systems to extract translation rules. This paper
presents the POS tagger which is also used as
segmenter, how to add POS to the phrase-based SMT
and the experiment. The experimental results showed
that the baseline SMT with POS that used POS
tagged language model could outperform the
baseline system in terms of BLEU and RIBES scores.
