A bloom filter based approach for evaluating structural similarity of XML documents

Abstract

Evaluating Similar structure of XML is a key issue for building the core algorithms for XML document clustering, XML classification and the extraction of schema or DTD from a corpus of XML documents. This evaluation is based on the structural similarity between XML documents. This work employs Bloom filter to represent an XML document with two structures: one is Tag-based Bloom filter (TBF) which describes an XML document with the tags of elements, and the other is Path-based Bloom filter (PBF) which describes hierarchical structure of the XML document. Based on this two structures, an approach is developed to evaluate the similarity of XML documents. A group of experiments was conducted to investigate the performance of the proposed approach. © 2009 Springer-Verlag.