现代藏文音节字自动校对研究

摘要

在现代藏文自动校对中,对音节字((阐)(鼬))的校对是其基础.现代藏文二维的书写格式和独特的文法,还有格助词的黏着现象、音节字搭配规则和音节字中真词和非词错误等众多问题,使得对藏文自动校对的研究有别于英语和汉语的自动校对.针对现代藏文中音节字的特点,通过音节字预处理、字表匹配、混淆集匹配、二元接续关系、最小编辑距离法等方法对现代藏文音节字的自动校对进行详细论述.%The syllables word proofreading is basic in the modern Tibetan automatically proofreading. The modern Tibetan has a lot of errors of two-dimensional writing format, unique grammar, the case auxiliary word bond phenomenon, syllables words connect rules, and the really-word and no-word in the syllables word etc that conduce to the Tibetan automatically proofreading difference between the English and Chinese. According to the characters of modern Tibetan syllables word, it uses the word syllables word pretreatment, word table matching, mixture set matching, the duality relationship, minimum edit distance etc to discuss the modern Tibetan syllables word automatically proofreading in detail.