International Journal of Computational Linguistics Research Chunye Zhang, Tianyue Yu, Yingqi Gao 16 1 2025 https://doi.org/10.6025/ijclr/2025/16/1/10-17 https://www.dline.info/jcl/fulltext/v16n1/jclv16n1_2.pdf Under the current wave of artificial intelligence, machine translation is a research direction in natural language processing, which has significant scientific and practical value. By analyzing the source language text in depth, we can utilize precise language features such as speech, vocabulary, intonation, etc., to transform it into understandable language results. Additionally, to address the machine translation task with a massive amount of language samples, the advantages of annotated data can be leveraged by combining the characteristics of neural networks to construct a more accurate translation model. Finally, by applying transfer learning, model overfitting can be avoided, greatly enhancing its generalization performance in less external input environments.