Elimination of Machine Translation Errors in English Language Transformation

Abstract

To improve the level of automation and intelligence of English language transformation in machine translation, a method of machine translation error elimination based on deep learning and feature extraction of language transformation error is proposed. The semantic correlation detection model of error exclusion in English language conversion translation is constructed by using the differentiated semantic modification method, and the semantic tree of error exclusion in English language transformation translation is built by means of grammar analysis. The semantic similarity feature of English language transformation is extracted. According to the different combinations of semantic similarity, the semantic allocation and machine translation error feature analysis in English language transformation are carried out. The tree topic word list of English language conversion is established by means of deep learning method, and the sentence structure of English language transformation is adjusted according to the semantic modification target in the tree topic word list. In order to eliminate the errors in translation of English language conversion and the registration of topic words, the optimal semantic correlation feature of each clause is calculated, and the deep learning algorithm is used to automatically optimize the errors in translation of English language conversion. The simulation results show that the accuracy of the proposed approach is high and the relevance of translation calibration is strong.