藏文组字部件的自动识别与字排序研究

Research on the Automatic Recognition and Sorting of Tibetan Word Components on the Unicode

藏文字有着独特的构字规则,组字部件的自动识别在藏文字、词、句层面的信息化处理有重要的应用。文章提出的藏文字部件自动识别方法理念为,现代藏文字按其第一个部件字符的不同分成5种结构类型,每一种类型又按其字长分为若干个子类,在每个子类中定义各字的部件识别算法,最后在藏文组字部件识别的基础上,给每个部件赋予序值,实现藏文字的有效排序。

Tibetan words have unique structure rules.The automatic recognition and sorting of word componentshas an important application in the information processing of various Tibetan word components such as charac-ter,word and sentence-level on the Unicode.In the present paper,according to the first component of Tibetancharacter,the Tibetan word structure can be divided into 5 categories.Each of word structure has been dividedinto several subcategories by the length of words.A recognition algorithm was defined for each word componentsin each subcategory.The ordinal value was given to each component of character based on the word componentrecognition to realizes sorting of Tibetan word efficiently.