基于编辑距离的XML文档结构聚类的改进算法  

Improved Algorithm of XML Document Structural Clustering Based on Edit Distance

作　　者：龚安[1] 刘华山[1] 

机构地区：[1]中国石油大学计算机与通信工程学院,东营257061

出　　处：《微计算机应用》2008年第2期88-91,共4页Microcomputer Applications

摘　　要：针对当前XML文档结构聚类算法的一些不足,指出XML文档树中节点的重复和嵌套影响聚类的质量和效率。利用重复剪枝和嵌套剪枝简化XML文档树的表示,然后根据化简后的结构计算两棵XML文档树中的编辑距离,在此基础上得出两棵树整体的结构相似度量,按照层次聚类方法得到聚类结果。实验证明该算法有比较高的查全率和查准率,有效降低了时间复杂性,具有改进效果。Because of node＇s repetition and nesting ,the performance and the quality of existed clustering algorithms is affected seriously, This article proposed a improved way for clustcring XML documents against the defects of the methods in existence. Firstly XML document tree is simplified by use of repetition reduction and nesting reduction ,then,we calculate the eidt distance according to simplified structure and structure similarity , finally get clustering result in terms of hierarchical methods. Experimental results prove that the algorithm has fairly high performance and can save time effectively.

关 键 词：XML 编辑距离 结构聚类 改进算法 

分 类 号：TP312.2[自动化与计算机技术—计算机软件与理论]