Design of Translation Accuracy Correction Algorithm for English Translation Software Based on Deep Learning

In today’s information society, how to quickly and efficiently break down language barriers between people through various technological means has become a common problem. Building the corpus required for statistical machine translation is an important step in ensuring the feasibility of this method in practical applications. Based on the initial language knowledge provided by linguists, semi-automatic processing of human-computer interaction is carried out, marking the part of speech, the meaning of each word in each phrase, the syntactic structure and semantic combination relationship between words, and turning it into a familiar corpus. In English translation, due to its own characteristics, people have conducted extensive research and analysis on it. In view of this, this paper makes a detailed study on the development and application of English translation software, analyzes sentences from bottom to top by using the moving-in convention algorithm, and processes tenses. Finally, the translated sentences are connected and the mood is treated accordingly.