古代汉字文献切分研究

针对古代汉字文档的特点,提出了适合于古文档的列切分方法和字切分方法。提出的列切分方法直接对文档的笔画投影进行分析,采用一种基于分层投影过滤和变长间隙阈值的递归切分算法。该算法在列间隔较小、列与格线存在粘连、文档具有一定程度的倾斜的情况下,也能准确地抽取出列,尤其对短列的切分达到了较好的效果。提出的字切分方法分为两步,进行粗切分确定大致的切分位置,采用基于连通域分析与粘连点判断的方法做进一步的细切分。该算法对具有较多粘连和重叠汉字的列,也能较好地切分出完整的单字。实验结果表明,提出的方法用于古代汉字文档切分能够获得较好的效果。

基于遗传算法的分类规则序列生成

与当前常用的分类方法相比,遗传算法具有较强的伸缩性和全局搜索能力,易于并行计算等优点.但通过遗传算法得到的一组分类规则之间常常存在冲突.本文先将分类规则表示成二进制编码,采用F-measure作为适应度评估函数,并设计了有效的杂交,变异等遗传算子,使遗传算法适合用在分类规则挖掘中.在遗传算法中增加了冲突解决机制,并结合顺序覆盖算法,使之可以得到分类规则的序列,解决了规则间的冲突,形成了一个完整的分类方法.最后针对具体实例作了测试,并将实验结果与分类算法J4.8得到的结果进行了对比,表明该方法略优于J4.8.

Interpretation Method of Guqin (Chinese Ancient Zither) Notation Based on Radical and Structural Analysis

Guqin music is a precious cultural heritage of China. The notation of Guqin is very special, which records its playing methods and techniques. For the purpose of preserving the guqin art, the digitalization of guqin notation and an interpretation method of guqin notation were conducted. By using this interpretation method, raw images of handwritten notations are transformed into structural data that can be processed and analyzed by computers easily. The method decomposes each single complex character of guqin notations into simple radicals and finds the structure of the character. According to the radicals and the structure, the character is interpreted into meaningful codes. The experimental results show our method is effective.