Clustering XML Documents by Structure

摘要

Clustering of XML documents is an important data mining method, the aim of which is the grouping of similar XML documents. The issue of clustering XML documents by structure is being considered in this paper. Two different and independent methods of clustering XML documents by structure are being proposed. The first method represents a set of XML documents as a set of labels. The second method introduces a new representation of a set of XML documents, which is called the SuperTree. In this paper, it is suggested that the proposed methods may improve the accuracy of XML clustering by structure. Such thesis is based on the tests, the aim of which is to assess advantages of the proposals, as conducted respectively on the heterogeneous and homogenous sets of data.