摘要
本文分析了单字分类的缺点 ,给出了以多字词及短语为分类的基本单位 ,利用MarKov模型、语言的多种统计知识及距离测度 ,进行分类的策略与方法。较好地克服了过去以单字为分类的基本单位造成的识别率较低、识别速度较慢的缺点。
This paper analyzes the shortcomings in the recognition of single Chinese characters and presents an approach in which a multicharacter word or a phrase is regarded as the basic recognition unit and the Markov model, statistical knowledge of the language and distance measurement are put to use so that it is possible to overcome the shortcomings of the low recognition rate and the slow recognition speed due to the treatment of a single character as the basic recognition unit
出处
《信息工程大学学报》
2000年第4期115-117,共3页
Journal of Information Engineering University
关键词
汉字识别
语言知识
Chinese character recognition
linguistic knowledge
