On Methods of Chinese Automatic Segmentation
Jie Chunyu, Liu Yuan, Liang Nanyuan
Automatic segmentation of character string into words is now referred to as another bottle-neck problem after Chinese character code in the field of Chinese information processing, On the base of reviewing and analyzing the previous methods of Chinese automatic segmentation, this article established a structure model ASM(d,a,m)to represent all basic methods systematically. And with this model,two new kinds of basic meshods were put forth. Furthermore,the calculation was made on the time complexity of each basic method;the influence of time complexity upon segmentation speed,and that of each basic method upon segmentation accurracacy and intelligent processing were analyzed in detail. Some wrong points of view on the methods of Chinese automatic segmentation were criticized as well.