Selection Method of Fuzzy Semantics in Machine Translation and the Integration of LBP Algorithm

Abstract

This paper studies the accuracy and rationality of machine English translation based on the LBP algorithm, and proposes a machine English translation method based on the selection of the optimal solution of fuzzy semantics. Construct an information extraction model for machine English translation, establish a fuzzy semantic topic word attribute table for machine English translation, and use phrases as the basic granularity to produce paraphrase results that are semantically consistent with the translation hypothesis set. Extract phrase paraphrase resources by using massively parallel corpus. Experimental test results show that using this method for machine English translation improves the recall performance of semantic information by 6.7%, and the feature matching degree of topic words is higher.