Research And Implementation Of Modify Chinese Part-of-Speech Tagging Based On FST Technology

Natural Language Understanding is also known as natural language processing or computational linguistics, it is one of the forefronts of problems in the field of artificial intelligence. Natural language recognition and processing is one of the most important topics in artificial intelligence research and is also the key to artificial intelligence research. Chinese Part-of-Speech Tagging is a fundamental subject to Chinese information processing technology in natural language processing; a precise part of speech tagging has a very wide range of meaning for accurate understanding of natural language, POS tagging is an essential task especially in syntactic analysis, semantic analysis. Therefore, research and implementation of Chinese tagging device is of great importance both in theoretical and practical aspect.There are two kinds of method in speech tagging, one is based on rules and another is based on statistics. Generally, in order to achieve better results of speech tagging, we often combine these two methods in practical application. Based on statistical methods, Hidden Markov Model (HMM) is mainly taken by, and we take Finite-State-Transducer (FST) approach at rule-based approach. So far, the theory reservation in the application of Natural language processing is deficient. In this paper I have done an further study on how to apply FST to natural language processing of speech tagging and given the results achieved ultimately.In recent years, under the influence of a new generation of computers in the fierce international competition, the study of this field has caused more and more attention. Research units and research teams are also gradually expanding. Currently in China, machine translation, corpus research, understanding research chapter and restricted Chinese studies are on behalf of the main results of the research. However, all these researches must have the front-end research on Part-of-Speech tagging.