DOI: https://doi.org/10.21203/rs.3.rs-2729745/v1
The constant emergence and rapid popularization of various intelligent technologies have brought a lot of convenience to people's lives, and also changed people's usual way of life. The use of machine automatic translation technology can greatly improve the efficiency of the analysis of publicity text information, and it is very helpful for people to deal with publicity text. The emergence of text machine automatic translation technology has brought convenience and new ideas to people's processing of large amounts of data. In the process of application, this technology will first model and analyze the semantic information contained in the text to be processed, and then output the information that people need according to their data processing requirements. In order to more clearly illustrate the effect of automatic text machine translation technology in practical applications, this paper selects two different types of text models, compares and analyzes the actual performance of this technology, and conducts a comparative study on the effect of Seq2Seq model and pre training model in translating text information. Combined with the relevant theory of deep learning, this paper illustrates the advantages and differences of the two models in translation effects, It provides scientific reference for the improvement of automatic translation model of publicity texts.