AUTHORSHIP ATTRIBUTION FOR CHINESE TEXT BASED ON SENTENCE RHYTHM FEATURES

	AUTHORSHIP ATTRIBUTION FOR CHINESE TEXT BASED ON SENTENCE RHYTHM FEATURES
	Wang SK(王少康)
	2010
会议名称	《科学数据库与信息技术》第九届学术研讨会
会议日期	2010
中文摘要	Authorship attribution, i.e., identifying the authorship of a piece of disputed text, is an important problem due to the increased concerns on copyright violations. While various authorship attribution algorithms have been proposed to identify the authorship of articles, they fail in several situations. This paper proposes a new authorship attribution algorithm for Chinese text using the sentence rhythm features of articles. In our algorithm, a rhythm feature matrix is proposed to depict the sentence rhythm of Chinese text. In order to determine the similarity of rhythm feature matrices, we compare two definitions of similarity based on Euclidean distance and improved Kullback-Leibler Divergence, respectively. Experimental results show that our algorithm achieves a success rate of 80%.
会议录	The 3rd IEEE International Conference on Advanced Computer Control
学科主题	算法理论
语种	中文
内容类型	会议论文
源URL	[http://ircnic.ac.cn/handle/311056/1811]
专题	计算机网络信息中心_中国科学院计算机网络信息中心(2012年前)_会议论文
推荐引用方式 GB/T 7714	Wang SK. AUTHORSHIP ATTRIBUTION FOR CHINESE TEXT BASED ON SENTENCE RHYTHM FEATURES[C]. 见:《科学数据库与信息技术》第九届学术研讨会. 2010.