GB18030汉字的分形相关性研究  

Study on the Fractal Correlation of GB18030 Words

作　　者：毛明毅[1] 陈志成[2] 李文正[1] 何华灿[3] 

机构地区：[1]北京工商大学计算机学院,北京100037 [2]清华大学信息技术研究院,北京100084 [3]西北工业大学计算机学院,西安710072

出　　处：《计算机科学》2007年第11期183-184,207,共3页Computer Science

基　　金：国家自然科学基金(编号:60273087)

摘　　要：GB18030是国家标准局新近颁布的最重要的汉字编码标准。本文从分形信息学的角度对GB18030汉字库中的27538个汉字的分形特性进行了研究。基于格分维理论,给出了汉字格测度的选取原则和格分维的计算方法。计算与统计表明:96.6846%的汉字的分形维数在[1.00,1.50]之间,98.9469%汉字的分形相关性系数R^2值在[0. 95,1.00]之间,这表明汉字具有显著的分形特性。GB18030 is the latest standard for encoding of Chinese words, which is constituted by the department of national standard of China. From the aspect of fractal informatics, this paper firstly studied the fractal charaeteristic of the 27538 Chinese words of GB18030 font libraries systemically. Based on the grid fractal theory, we give the principles of selecting grid measure for Chinese words and the procedure of calculating grid fractal dimension. The result is that and the grid fractal dimension of 96. 6846% words are in range[1.00, 1.50], and the fractal corrdation coefficient R^2 of 96. 6846% words are in range[0. 95, 1. 00]. This shows that the Chinese words have notable fractal characteristic.

关 键 词：GB18030 汉字 格分维 相关性 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]