节点文献
现代汉语诗歌与其他文体文章的字同现网络之比较
Comparison of co-occurrence networks of Chinese characters:from poems to articles in modern Chinese language
【Author】 Wang Yanli1,Shi Yuming1,Liang Wei1,G uanrong Chen2 1Shandong University,Jinan 250100,China 2City University of Hong Kong,Hong Kong,China
【摘要】 现代中文诗歌主要分三类:自由诗、格律诗、散文诗。我们分别建立了此三类现代汉语诗歌的字同现网络,并研究了它们的统计特征:度分布、聚类系数、平均最短路径及其连通性等。在字同现网络中,每个字用一个节点来表示;两个字如果出现在同一个句子中且相邻,则用一条边把它们相连。我们发现,所研究的中文诗歌的字同现网络都具有无标度特性及小世界特性;但是,它们的聚类系数、平均最短路径和幂率指数之间存在一些差别。我们进一步比较了现代汉语诗歌与其他文体文章(散文、小说、科普、新闻)字同现网络的统计特征,发现不同文体之间也存在一些差别。本项研究从复杂网络的角度揭示了不同文体之间、以及三类诗体之间的异同。
【Abstract】 Modern Chinese poems may be divided into three categories according to their styles:free verses,regulated verses,and prose poems.Co-occurrence networks of Chinese characters are constructed and compared for each poetry style in terms of their statistical parameters,including degree distribution,clustering coefficient,average shortest path length,and connectedness.In the constructed networks,a character is represented by a node and two characters are connected by an edge if they are adjacent in the same sentence.All the co-occurrence networks of Chinese characters that have been studied for different styles of poetries exhibit small-world and scale-free features.However,there are some differences among their clustering coefficients,average shortest path lengths,and power-law exponents.In addition,a comparison of the statistical parameters of networks between poems and articles of the other types(essays,novels,popular science articles,and news reports) in the modern Chinese language also shows some differences.The present study reveals,from a network perspective,some similarities as well as differences among three styles of poetries and five types of articles in Chinese language.
【Key words】 Chinese poem; Co-occurrence network of Chinese characters; Small-word network; Scale-free network;
- 【会议录名称】 第五届全国复杂网络学术会议论文(摘要)汇集
- 【会议名称】第五届全国复杂网络学术会议
- 【会议时间】2009-10-15
- 【会议地点】中国山东青岛
- 【分类号】I207.25
- 【主办单位】青岛大学、中国工业与应用数学学会