Tibetan-Chinese Machine Translation Technology Combining Linguistic Grammar and Formal Grammar

Abstract

Open and closed experiments were carried out on the corpus respectively. Machine translation (MT) model based on syntactic information is better than that based on phrase model, which proves that adding Tibetan short syntactic analysis to Tibetan-Chinese MT application system is helpful to improve the performance of Tibetan-Chinese MT system(TCMT). What’s more, it also proves that the ITG is suitable for TCMT and can tackle the problem of inconsistent the order of the words between two languages in machine translation, which improves the performance of the overall TCMT system