XQuery speedup by deploying structural redundancy in mapping XML into relations

Abstract

In designing a relational schema, we often consider that an attribute of a table is replicated into other table to reduce the join cost. Maybe such a possible redundancy will be grasped through E/R model (i.e. semantic analysis). Similarly, in mapping XML into relations, we can consider some redundancies to enhance query performance and they can be grasped through the structural traits of DTD (or XML schema). Several practical structural redundancies are formulated in this paper. If given XML data and queries are very large and complex, finding essential replications may also be difficult, and two efficient search methods are introduced for helping the search. Since the search problem is NP-hard, the methods are heuristically designed. Finally, read and update query costs arising by employing the structural redundancy are analyzed experimentally and the efficiency of two search methods is analyzed. They showed that the replication strategy can be very useful.