Construction and Optimization of Computer Intelligent Recognition and Analysis System Based on Natural Language Processing

Abstract

Applied researches with computer intelligent recognition technology as the core have gained a rapid growth, which are based on natural language processing. The constantly acquired practical applications make researches on natural language processing based on computer intelligent recognition technology become a key field of automatic speech recognition (ASR). The establishment of a hierarchical structure for an English translation recognition system includes a translation material collecting module, an information feature extraction module, an analysis model construction module and a result feedback grading module; establish a linguistic model for English translation recognition systems, with which the statistics of probability distribution of certain sentence or word sequences in the translation can be completed, and the information characteristics of English documents translated by users as well as translation training sets are extracted; calculate the similarity among characteristic keywords based on the feature extraction result, conduct fitting calculations using a BP network after particle swarm optimization, and finally realize the establishment of an intelligent recognition system for English translation.