【摘要】 中文内容识别是目前研究界关注的热点,目前对非正常书写的汉字词组识别的质量和速度不能满足应用需求.为解决上述问题,本文提出了基于音律的中文内容识别的新方法,主要工作包括:①分析现有方法的不足之处;②提出了汉字热码的概念和基于热码的内容识别的方法;③提出并实现了音律首字母树结构和基于首字母树的算法;④做了详实的实验,实验表明,新的中文内容识别方法准确度提高了18%,速度较传统方法提高了4倍.更多还原

【Abstract】 Identify Chinese content attracts dedicated studying both in research fields and industry.Current techniques cannot meet the requirement of applications in both quality and speed when meeting with irregular written phrases.In order to solve this problem,a new method named Phonetics-Code-Identification(PCI) is proposed.The contributions includes : (1) Analyzes the weakness of traditional methods.(2) Proposes the conception of Chinese HotCode and the methods of content identification based on HotCode.(3) Suggests a structure named First-Character-Tree(FCT).(4) Experiments showed that the new method improves the identification accuracy by 18%,increases the speed by 4 times compared with traditional methods.更多还原