Machine Automatic Translation Quality Evaluation Model Based on Recurrent Neural Network Algorithm

Abstract

AbstractIn real life, the evaluation of translation quality is widely used. Not only translation activities need evaluation and comparison, but also translation quality evaluation in the editing and publishing of foreign language works, text translation teaching and other fields. The purpose of this paper is to study the quality evaluation model of machine automatic translation based on the recurrent neural network algorithm. First introduced the research background of this article, discussed the importance of research, and then briefly reviewed the research and development of machine translation evaluation, focusing on the current situation of recurrent neural network and evaluation method automation, and summarized the translation evaluation of automated machine technology. Finally, the main content of this article is summarized. The language model based on iterative neural network algorithm is studied. Basic experiments are carried out on the repetitive neural network translation quality evaluation model for machine translation quality evaluation. On the basis of basic experiments, 100 sentences were randomly selected to evaluate the performance of the translation system. The experimental results show that the translation quality estimation system of the machine automatic translation quality evaluation model based on the recurrent neural network algorithm established in this paper is 0.1 points better than the QuEst system.KeywordsRecurrent neural networkMachine translationAutomatic translationQuality assessment