Research on Bilingual Dong Cuisine Translation System Based on Natural Language Processing

The research on bilingual Dong Cuisine Translation System based on natural language processing aims to solve the language barrier problem in the dissemination of Dong cuisine culture. This study combines natural language processing technology and GPT model to design and implement an efficient bilingual Dong Cuisine Translation System. The core algorithms of the system include text preprocessing, syntactic analysis, and Dong-Chinese bilingual conversion based on GPT model. To improve the accuracy of translation, this study conducted an in-depth analysis of the syntactic structure of Dong language and English, and designed an optimization algorithm adapted to Dong language grammar. Many Dong cuisine menu data sets were used in the model training process. Through supervised learning and multiple rounds of tuning, the model can accurately capture the subtle differences in Dong language. In the simulation test, the system used Dong cuisine recipe data from different fields for evaluation. The simulation results show that the system performs well in translation accuracy and fluency, with the translation accuracy of Dong language to English reaching 92.7% and the translation accuracy of English to Dong language reaching 90.3%. At the same time, through the analysis of indicators such as BLEU score, the system has strong robustness in cross-language translation tasks and can effectively reduce the ambiguity problem in translation.