CAJ | 학술논문

各类维吾尔文本信息处理技术需要准确、快速、高效的音节自动切分的支持。目前通过辅助音节库的方法解决基于非规范音节结构的外来词语的音节切分。本文在分析维吾尔文音节结构和切分规则基础上，用后序遍历、音节结构的检查，修复不规则音节的方法实现了不依赖附加音节库的全自动音节切分，提出人工生成有限数量的模拟词语的音节切分测试方法，分别用词典和基于模拟词语测试，切分准确率分别达到了100％和96％。
각류유오이문본신식처리기술수요준학、쾌속、고효적음절자동절분적지지。목전통과보조음절고적방법해결기우비규범음절결구적외래사어적음절절분。본문재분석유오이문음절결구화절분규칙기출상，용후서편력、음절결구적검사，수복불규칙음절적방법실현료불의뢰부가음절고적전자동음절절분，제출인공생성유한수량적모의사어적음절절분측시방법，분별용사전화기우모의사어측시，절분준학솔분별체도료100％화96％。
An accurate and efficient auto syllable segmentation is the basic technique in Uyghur information science.Additional ir-regular foreign words library is a present solution.According to the latest modern Uyghur syllable structure and man-made seg-mentation method,the post-order traversal,inspection and repair of syllable structures are applied to segment the syllable with-out additional libraries.A new syllable segmentation test method based on artificial created words is proposed.The test accuracy based on dictionary and artificial words is up to 100% and 96% respectively.