Statistic machine translation system based on syntax framework

Abstract

The invention relates to a statistic machine translation system based on a syntax framework. The translation process comprises the steps that 1, non-syntax translation rules are extracted through a probabilistic SCFG-level rule extraction method and used for translation of a non-framework part of a sentence to be translated; 2, syntax translation rules are extracted through a GHKM rule method and used for translation of a framework part of the sentence to be translated; 3, non-complete syntax translation rules are generated according to the syntax translation rules, and the non-syntax translation rules and the syntax translation rules are combined to realize integration of advantages of a non-syntax translation system and advantages of a syntax translation system; 4, a model is generated. According to the system, the syntax translation rules are used for translation of the syntax framework and long-distance sequencing, and the rules of the non-syntax translation system are used for processing low-level vocabulary translation and sequencing; the model is easy to realize, and the effect is remarkable.

Description

technical field [0001] The invention relates to a technique for modeling source syntax in statistical machine translation, in particular to a statistical machine translation system based on a syntactic skeleton. Background technique [0002] In Statistical Machine Translation (SMT), there are different translation systems, such as non-syntactic translation systems based on phrases and hierarchical phrases, and syntactic translation systems such as tree-to-string and string-to-tree. Different translation systems have their own advantages and disadvantages. For example, the syntactic translation system has obvious advantages in dealing with long-distance and complex ordering problems between various components, but when the translation rules of syntactic translation are relatively sparse or the coverage When it is relatively low, there will be system robustness problems, which may lead to poor translation effects. And it has been confirmed that if the syntactic system is impl...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problem of system robustness caused by the inability to translate and sequence short sentence fragments well and sparse rules in the syntactic translation system in the prior art, and the model cannot perform long-distance sentence components in the non-syntactic translation system The problem of effective ordering, time-consuming and labor-intensive human-labeled skeleton information, etc., the technical problem to be solved by the present invention is to provide a statistical machine translation system based on syntactic skeleton, which models the high-level syntactic skeleton of the source language, and good translation of low-level phrases while proposing a novel representation of the syntactic skeleton for machine translation systems to use

Images