Detecting Changes on Unordered XML Documents Using Relational Databases: A Schema-Conscious Approach

Abstract

Several relational approaches have been proposed to detect the changes to XML documents by using relational databases. These approaches store the XML documents in the rela-tional database and issue SQL queries (whenever appropriate) to detect the changes. All of these relational-based approaches use the schema-oblivious XML storage strategy for detecting the changes. However, there is growing evidence that schema-conscious storage approaches perform significantly better than schema-oblivious approaches as far as XML query processing is concerned. In this paper, we study a relational-based unordered XML change detection technique (called Helios) that uses a schema-conscious approach (Shared-Inlining) as the underlying storage strategy. Helios is up to 52 times faster than X-Diff [7] for large datasets (more than 1000 nodes). It is also up to 6.7 times faster than Xandy [4]. The result quality of deltas detected by Helios is comparable to the result quality of deltas detected by Xandy.