The development of machine translation has also experienced many twists and turns. A French engineer put forward machine translation in 1933, and obtained the patent of translation machine; Georgetown University completed the machine translation experiment for the first time in 1954. The American Academy of Sciences denied the feasibility of machine translation in 1966, and the research of machine translation entered a frustrating period. The University of Montreal in Canada and the Translation Bureau of the Canadian Federal Government jointly developed the TAUM-METEO system in 1976, which marked the revival of machine translation. Since then, machine translation has entered a stage of development and prosperity.
The preprocessing stage of machine translation is the regularization source. Turn a long sentence into several short sentences, delete unnecessary parts, and adjust the irregular expressions.
The core processing stage is the core step of machine translation. The core processing translates character units and source sequences into target sequences.
In the post-processing stage, the translation results are spliced and adjusted to make them conform to people's reading habits. These include: modeling unit splicing, letter case conversion, special symbol processing and so on. This is the general process of machine translation.
People's quality of life is constantly improving, and people travel everywhere. But people can't master every language, so machine translation comes in handy. It can make communication between people from different countries barrier-free and expand people's contact. Although there are still some problems in machine translation, with the development of science and technology, these problems will be solved.