Interactive machine translation method based on bilingual segments

Abstract

The invention relates to an interactive machine translation method based on bilingual segments. The method includes the steps of establishing a mathematical model, wherein for each source language segment, a translator is provided with multiple translation options, and the optimal translation is obtained through the mathematical model; designing a translator interface, wherein the translator interface includes an interaction area and an editing area, the interaction area gives a source sentence and translation options after phrase segmentation, and the editing area gives a machine translationwhen the translator completes the confirmation and clicks a 'Translate' button; decoding, wherein after the translator confirms the bilingual segments in the interaction area, the translator's choiceof translation options for each segment fi and the current segmentation result of the source sentence are acquired, and a phrase-based statistical machine translation decoder is achieved by a multi-stack decoding algorithm. The method improves interactive protocols, allows the translator to confirm the bilingual segments, provides more clues to the translator, and gives the decoder more direct guidance, reduces the human labor in a human-computer interaction process, and promotes the improvement of interactive machine translation efficiency and translation quality.

Description

technical field [0001] The invention relates to a natural language translation technology, in particular to an interactive machine translation method based on bilingual fragments. Background technique [0002] Statistical machine translation and neural machine translation techniques have greatly improved the performance of machine translation systems. However, in many high-quality tasks, the output quality of machine translation is still insufficient, and must be modified by human translators in the post-editing process before it can be used. [0003] In order to enhance human-computer cooperation, Foster proposed interactive machine translation technology. In an interactive machine translation system, a modification-prediction process is iteratively performed. First, the interactive machine translation system provides an initial translation. The translator then identifies the longest correct prefix among them and corrects the next word. Next, the system predicts a new s...

AI Technical Summary

Problems solved by technology

However, there are still three problems with this protocol: first, the positions of the confirmed fragments are unknown, so the search process can only be optimized in the form of a soft constraint; second, the translator's confirmation is limited to the translations provided by the system third, identifying the correct segment from the wrong translation often requires a lot of cognitive effort, especially when the translation is of low quality

Images