Exploring the Strategy of English Translation of Ancient Poems with the Assistance of Neural Network Model

Abstract

In the background of globalization, the cultural exchange between countries all over the world is a great trend, and thus the translation between different languages is a difficult problem. The ancient poems are the splendid treasures of our culture, but the translation of ancient poems is a big problem, and the manual translation is a troublesome task, so the means of machine translation is a good solution. Nowadays, the development of neural network is extremely fast, and it is perfectly feasible to translate ancient poems by the machine translation trained by neural network. In this paper, we improve the machine translation model based on recurrent neural network, obtain the data set required for training independently, and train the machine translation from it, and get significantly better results than the ordinary model in the English translation of ancient poems.