Research On Chinese-Vietnamese Phrase Machine Translation Method Based On Pivot Language

Vietnam,as a neighboring Southeast Asian country bordering on China,under the influence of the "One Belt and One Road" policy,has become increasingly close to China's political,economic,and cultural exchanges.The importance of Chinese and Vietnamese bilingual translation is increasingly strengthened.Machine translation is one of the important research areas of natural language processing and it is the current mainstream translation method.Therefore,the study of Chinese Vietnamese statistics Machine Translation has an important supporting role in the translation,information retrieval,cultural exchange and economic trade of the Chinese and Vietnamese bilingual translation.At present,the translation model of Chinese to Vietnamese is still in its infancy.Some main works focus on the construction of bilingual parallel corpus,the research of Chinese and Vietnamese word alignment,and the dependency syntax tree of Vietnamese.Because Vietnamese is a small language in Southeast Asia,there are few parallel corpus resources between Chinese and English,so it is difficult to get better model training effect.Therefore,it is the focus of this paper to introduce English as an intermediate language into Chinese Vietnamese bilingual translation.(1)Pivot phrases probability table based on Wikipedia.Since Vietnamese is a resource scarcity language,there are no Chinese Vietnamese sentences aligned at present,and Han Vietnamese phrases lack data support for extraction.Therefore,taking English as the pivot language,Chinese English,English and Vietnamese phrases are extracted.We collected the same entries from Wikipedia on the Chinese,English and Vietnamese parallel web page.By analyzing the structure of the web page and the vocabulary in the web page,the parallel sentences are extracted from the parallel pages.Finally,the construction of the pivot phrase probability table is completed.(2)The Chinese Vietnamese phrase translation method is combined with the pivot phrase probability table.The phrase probability table is the key to the training of the phrase translation model.At present,the size of the Han Vietnamese phrase probability table is small,and it can not effectively support the Chinese Vietnamese phrase translation.So we use the phrase translation probability and phrases in vocabulary of weights to establish a more Chinese bilingual phrase translation model bridge.Based on the phrasal Machine Translation model,the Han Yue phrase probability table is incorporated.The pivot phrase probability table is incorporated into the existing phrase translation system.The experimental results show that the method can effectively improve the accuracy of the translation system.(3)The Chinese Vietnamese phrase translation method is combined with the pivot phrase probability table.The phrase probability table is the key to the training of the phrase translation model.At present,the size of the Han Vietnamese phrase probability table is small,and it can not effectively support the Chinese Vietnamese phrase translation.So we use the phrase translation probability and phrases in vocabulary of weights to establish a more Chinese bilingual phrase translation model bridge.Based on the phrasal Machine Translation model,the Han Yue phrase probability table is incorporated.The pivot phrase probability table is incorporated into the existing phrase translation system.The experimental results show that the method can effectively improve the accuracy of the translation system.