Translation method and device in statistical machine translation

Abstract

The invention discloses a translation method and device in statistical machine translation. During a training stage of the statistical machine translation, a translation model which comprises the semantic vector of a source end phrase is generated, the semantic vector of a phrase to be translated is calculated, then, similarity between the semantic vector of the phrase to be translated and the semantic vector of the source end phrase of a candidate translation rule is calculated, and the similarity is integrated into a log-linear model as a submodel set for a statistical machine translation system to select a translation which is most similar to the semantics of the phrase to be translated. The traditional statistical machine translation system does not contain semantic information. The translation method and device disclosed by the invention generates the semantics of the source end phrase and utilizes the similarity of the semantic vector to help the statistical machine translation system to more accurately select the translation which is similar to the semantics of the phrase to be translated, and can effectively improve the quality of the translation of the statistical machine translation.

Description

technical field This application relates to the field of data processing, more specifically, to a translation method and device in statistical machine translation. Background technique With the development of the Internet and social networks, machine translation plays an increasingly prominent role in social development and information dissemination. Statistical Machine Translation (Statistical Machine Translation, SMT) is an important research direction of machine translation. Its basic idea is to construct a statistical translation model through statistical analysis of a large number of parallel corpora, and then use this model for translation. From early word-based machine translation to phrase-based translation, the quality of translation has been greatly improved. However, in statistical machine translation, a source phrase may correspond to multiple translation rules in the trained translation model. For example, the English phrase "runa" can mean "run one" or "run...

AI Technical Summary

Problems solved by technology

At present, the translation system in statistical machine translation only selects the translation with the highest probability from multiple candidate translation rules based on the vocabulary translation probability, which sometimes causes a large deviation between the translation of the entire sentence and the meaning expressed by the source sentence

Images