Optimization of Translation Techniques between English and Chinese Literary Works in the Information Age

Abstract

Under the background of cultural globalization, English-Chinese literary translation plays an important role in which it is not only a process of language conversion, but also a process of cultural conversion. In this paper, the BiGUR-LM-Attention optimization model is fused and constructed using the WordNet semantic similarity model, GRU-LM one-way gated similarity model, and BiGR-LM two-way gated similarity model. The LDA theme model is selected to generate the 3-layer Bayesian network structure of literary works’ paragraphs, themes and words to obtain the probability information that represents the highest attention of the work’s text theme, which constitutes the attention mechanism feature word vector. Finally, five classic literary works are selected as the training corpus to compare and analyze the translation quality between machine translation and human translation in the mutual translation of English and Chinese literary works. The results show that the number of errors and the total score of machine translation are 95 and 275, which are significantly lower than those of manual translation, 105.37 and 360.19. The new model has outstanding translation performance in semantic recognition, dialect, and special nouns, which effectively improves the translation quality of literary works and is of great significance for the dissemination of cultural works.