Chinese/Tibetan named entity inter-translation method and device

Abstract

The invention relates to a Chinese / Tibetan named entity inter-translation method and device. According to the method and device, a Chinese / Tibetan contrast named entity database is established, and aunique corresponding translation word is distributed for each of the different named entities, so that one-to-one correspondence between Chinese named entities and Tibetan named entities in Chinese / Tibetan inter-translation process is realized, and then the correctness of machine translation and the correctness during restoration are improved.

Description

technical field [0001] The invention relates to the technical field of language translation, in particular to a method and device for inter-translating Chinese-Tibetan named entities. Background technique [0002] With the development of computer network platforms, information transmission has become a very important content in people's life and work, and machine translation has also developed accordingly. However, the accuracy of translation has become a major bottleneck for people to widely use machine translation systems, especially in Chinese. There is no uniform translation standard for named entities in Tibetan machine translation. In addition, when ordinary Chinese characters are used for named entities, the phenomenon of different sounds, different words with the same pronunciation, and multiple parts of speech in a word is common. Named entities with different sounds, homonyms, and multiple parts of speech in a word are translated into the same Chinese character or ...

AI Technical Summary

Problems solved by technology

[0002] With the development of computer network platforms, information transmission has become a very important content in people's life and work, and machine translation has also developed accordingly. However, the accuracy of translation has become a major bottleneck for people to widely use machine translation systems, especially in Chinese. There is no uniform translation standard for named entities in Tibetan machine translation. In addition, when ordinary Chinese characters are used for named entities, the phenomenon of different sounds, different words with the same pronunciation, and multiple parts of speech in a word is common. Named entities with different sounds, homonyms, and multiple parts of speech in a word are translated into the same Chinese character or the same Tibetan word according to free translation or transliteration, which is very detrimental to the accuracy of machine translation, and when it is necessary to restore the translation to the source language When , it is almost impossible for existing translation methods to achieve accurate restoration

Images