Uygur language part-of-speech tagging method

Abstract

The invention discloses a Uygur language part-of-speech tagging method. The method includes 1, formulating a Uygur language part-of-speech tagging set and a million-word Uygur language corpus; 2, selecting a method based on conditional random fields in primary tagging to build a Uygur language part-of-speech tagging model, wherein the method is flexible in feature extraction and high in accuracy; 3, building a correct tagging rule library, an unambiguous part-of-speech tagging dictionary and a proper noun dictionary, and building a primary part-of-speech tagging correction algorithm based on rules and dictionaries to further improve accuracy of primary part-of-speech tagging; 4, providing a part-of-speech tagging method based on stem extraction to further increase coverage rate of tagged words; 5, providing a secondary part-of-speech tagging statistical model to increase coverage rate and success rate of the tagged words; 6, tagging in secondary tagging through the unambiguous dictionary and the proper noun dictionary, and realizing secondary part-of-speech tagging with extremely high accuracy through stem extraction tagging and statistical model tagging. By the Uygur language part-of-speech tagging method, the problem of part-of-speech tagging of Uygur language is solved efficiently.

Description

technical field [0001] The invention relates to a language information processing technology, in particular to a part-of-speech tagging method of Uighur language. Background technique [0002] Today, with the informatization of the national economy and society, massive amounts of information are generated, stored, and disseminated every day. Human beings are faced with unprecedented information expansion. How to find the information they need from the massive amount of information The information that can be understood by oneself has become a common concern of people, and it is also a problem that information processing needs to solve. At present, natural language processing has become a striking research hotspot in the field of information processing. [0003] The Xinjiang Uygur Autonomous Region is a region where many ethnic groups live in concentrated communities. Among the current population of more than 20 million, there are more than 13 million ethnic minorities, acco...

AI Technical Summary

Problems solved by technology

A large number of scholars at home and abroad have conducted a lot of in-depth and detailed research on English and Chinese, but the research on Uyghur is basically blank

Images