Phrase Based Language Model For Statistical Machine Translation
This work addresses a specific bottleneck in machine translation by adapting phrase-based models from speech recognition, representing an incremental advancement.
The paper tackles the problem of improving statistical machine translation by introducing phrase-based language models, which outperform word-based models in perplexity and translation quality.
We consider phrase based Language Models (LM), which generalize the commonly used word level models. Similar concept on phrase based LMs appears in speech recognition, which is rather specialized and thus less suitable for machine translation (MT). In contrast to the dependency LM, we first introduce the exhaustive phrase-based LMs tailored for MT use. Preliminary experimental results show that our approach outperform word based LMs with the respect to perplexity and translation quality.