A Multilingual Professional Translation Method Based on Sentence Components

Abstract

The invention relates to a sentence component device and a multi-language professional translation method based on the sentence component. The device comprises a sentence pattern bank, a cabin model bank, a sense-group serial bank and an idiom bank; the sentence components in the bank appear as the semantic blocks of different sizes, comprising both frameworks and components. The multi-language specialized translation method based on the sentence component can automatically or semi-automatically provide the target sentences according to different support degrees of four kinds of banks and the method can provide the target sentences through organically combining the prior art based on the machine-translation assistance of sentences against a memory bank. With the expanded translation memory, 'the same sentences does not need to be translated for the second time', and 'the sentences of the same sentence pattern can be automatically or semi-automatically provided with the machine translation of consistent meaning'. Therefore, the translation automation level is enhanced and the requirement on the foreign language level of operators can be reduced. During the professional translating process and both before and after the translating process, the bank can be automatically or semi-automatically expanded; and the more the device and the method are applied, the better the application effect is.

Description

【Technical field】 [0001] The invention relates to a device and method for processing or converting natural language; in particular, it relates to a multilingual machine-aided translation method used in the professional field. 【Background technique】 [0002] Machine translation, such as the book "Computer Processing of Natural Language", written by Feng Zhiwei, published by Shanghai Foreign Language Education Press in October 1996, the first section of the eighth chapter is machine translation. The full text expounds in detail that in the early 1930s, the French scientist Aljuni proposed the idea of ​​using machines for language translation; by 1946, when the world's first computer came out, the research on machine translation began in the same year, and it flourished for a while; 1966 In November 2010, the Automatic Language Processing Advisory Committee of the American Academy of Sciences published a report entitled "Language and Machines (ALPAC)", which denied machine tran...

AI Technical Summary

Problems solved by technology

"Finally, the author pointed out: "The problem of machine translation technology itself is the flaw that hinders its development.

Machine translation is based on rules, limited by the level of artificial intelligence, it cannot adapt to the particularity and diversity of languages, and the translation quality is too poor

Memory translation is based on the memory of source and target sentence pairs, and the translation quality can reach professional level, but it is only suitable for a small part of repetitive texts, and it is powerless for most non-repetitive texts

Images