Functional Noun Phrase Recognition in English Machine Translation Based on Machine Learning

With the development of science, the recognition of English translation noun phrases in machine learning plays an important role in machine translation. One of the bottlenecks in current English machine translation is the ambiguity resolution of noun phrases. This paper proposes a neural network machine translation functional phrase and its recognition method to eliminate noun phrase structures in English machine translation. The boundary of a noun phrase is determined based on the functional grammar of machine learning in clauses. Such noun phrases are referred to as functional noun phrases in this article. The research work mainly includes the analysis of incorrect translation of noun phrases in English machine translation, and the beneficial comparison between machine translation and human translation. The experiment of constructing a phrase corpus in this paper analyzes a small-scale phrase corpus with a tag size of about 580000 English words, which can construct approximately 20000 sentences and 10000 phrases, while a regular phrase corpus can only construct 10000 sentences.