基于聚类技术的XML文件代表性结构获取

XML文件可以利用树状结构来表示，于是把如何将XML文件做聚类看成如何对树状结构的数据作聚类.使用SOM聚类工具搭配上Jaccard 的距离测量公式来对XML 文件做聚类，然后在每个cluster 中利用GST(Graph Search Technique)算法从这些XML文件当中找出他们的最大序列，最后将这些最大序列融合起来成为共同的结构.

Representative Structures from XML Documents Based on Clustering Techniques

Since an XML document can be represented as a tree structure,the problem how to cluster a collection of XML documents can be considered as how to cluster a collection of tree-structured documents.The author used SOM (Self-Organizing Map) with the Jaccard coefficient to cluster XML documents.Then,an efficient sequential mining method called GST was applied to find maximum frequent sequences.Finally,the author merged the maximum frequent sequences to produce the common structures in a cluster.