一种适用于唐诗诗句分词方法的研究  

Research on Segmentation Methods Applicable to Tang Poetry

作　　者：闫伟[1] 杨秀梅[2] 

机构地区：[1]四川大学计算机学院,成都610000 [2]中国科学院大学,北京100049

出　　处：《现代计算机》2016年第2期17-19,共3页Modern Computer

摘　　要：构建一个全唐诗语料库是使用计算机自动研究唐诗的基础。在对唐诗中的"词"、"复合词"等进行系统定义的基础上,综合唐诗在格律方面的特点,提出一种适用于唐诗诗句分词的方法。该方法先依据唐诗格律特点对诗句进行切分,对不符合词定义的切分结果,再采用现代汉语分词技术中最大逆向匹配算法进行修正。实验结果表明,所提出的方法能有效提高唐诗分词准确率。The construction of a full Tang poetry corpus is the basis for the study of Tang poetry by computer. Based on the Tang Dynasty in the ＂word＂ and ＂compound＂ system is defined on the basis of the comprehensive characteristics of Tang poetry in metrical form, proposes a method applicable to segment the Tang poetry. The first method of Tang poetry metrical features segmentation of verse basis, does not conform to the definition of the segmentation results, and then the use of modem Chinese word segmentation in reverse matching algorithm is modified. The experimental results show that the method proposed can effectively improve the accuracy rate of Tang poetry word segmentation.

关 键 词：唐诗 语料库 分词技术 字串匹配 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]