书面藏语的续连规则及词形变体的自动识别  

The Machine Identification of Affixing-rules in Writing Tibetan.

作　　者：江荻[1] 

机构地区：[1]中国社会科学院民族学与人类学研究所,北京100081

出　　处：《民族语文》2004年第6期50-54,共5页Minority Languages of China

基　　金：国家自然科学基金资助(项目批准号为60173024)。

摘　　要：本文是现代书面藏语不同类型句法单位变体形式的机器自动识别研究,包括词形 变化的单位分类,抽取变体形式的信息和对文本进行句法语义功能的标注处理方法。This paper discusses how to build up machine identification rules for differentiating the heterographs of Tibetan words and markers on the basis of affixing-rules database. The Tibetan affixing-rules database contains more than ten types of words, which change forms in texts according to the pre-syllabic ending forms. So the paper first describes the types of heterographic variants and their conditions, then advances a method to extract information of variants and goes on tagging and processing all the heterographs in the text.

关 键 词：书面藏语 变体 词形变化 句法语义 对文 形式 单位 不同类型 信息 规则 

分 类 号：H214[语言文字—少数民族语言]