Applications Research of Machine Learning Algorithm in Translation System

In recent years, machine translation has made outstanding achievements, but there are also the problems such as fuzzy rules, insufficient data and improper order in the machine translation practice, which makes it still unsatisfactory. Machine learning can abstract learning features, and establish complex mapping relationship between input and output signals, and effectively improve the problems existing in the translation system. The concepts and common models of machine learning and deep learning are given in this paper. Based on the neural model, the paper analyses the rules, structure and models of machine translation system. According to the requirement of the translation system, this paper develops a translation system based on machine learning, and gives the design process of the pre-processing module, the coding module, the attention module and the decoding module. Practice has proved that the system has a good performance in translation performance.