Correction of English Translation Accuracy Based on Poisson Log-linear Model

Abstract

Abstract In the conventional machine translation methods, pipelined sequential operation is used to perform part-of-speech identification and syntactic analysis on the raw corpus to obtain the syntactic structure of English language, thereby reducing the iterative transfer error between translation tasks and the accuracy of structured instances, resulting in reduced accuracy in English language and literature translation. In this paper, a Poisson log-linear model that saves the corresponding bilingual corpus by means of Chinese-English dependency-tree-to-string is designed to implement dependent structured processing on the source language end and ensure that the accurate translation of English language is further proofread through the data-oriented translation model in the Chinese-English bilingual correspondence. The experimental results show that translation with high accuracy can be obtained based on the proposed method, which is highly accurate and stable.