An XML Index on B + -Tree for Content and Structural Search
Toshiyuki Shimizu, Masatoshi Yoshikawa · 2005
XML query processing is one of the most active database research areas. Although the main focus of the past research has been the processing of structural XML queries, there are growing demands for full-text search for XML documents. In this paper, we propose new indices which aim for high-speed processing of both full-text and structural queries on XML documents. An important design principle of our indices is the use of B-tree. To represent structural information of XML trees, each node in the XML tree is labeled an identifier. The identifier contains an integer number representing the path information from the root node. We have designed two types of indices, STB-tree and COB-tree using B-tree. Index entries of the COB-tree are a pair of text fragment in the XML document and the identifier of the leaf node which contains the text, whereas index entries of the STB-tree are an identifier of nodes. We have implemented COB-tree and STB-tree using GiST. We will show the efficiency of our indices in the experimental study.