Metamodel-Based Optimisation of XPath Queries

  • Authors:

  • Gerard Marks;Mark Roantree

  • Affiliations:

  • Interoperable Systems Group, Dublin City University, Ireland;Interoperable Systems Group, Dublin City University, Ireland

  • Venue:
  • BNCOD 26 Proceedings of the 26th British National Conference on Databases: Dataspace: The Final Frontier
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

To date, query performance in XML databases remains a difficult problem. XML documents are often very large making fast access to nodes within document trees cumbersome for query processors. Many research teams have addressed this issue with efficient algorithms and powerful indexes, but XML systems still cannot perform at the same level as relational databases. In this paper, we present a metamodel, which enables us to efficiently solve relationships between nodes in an XML database using standard SQL. By implementing the metamodel presented here, one can turn any off-the-shelf relational database into a high performance XPath processor. We will demonstrate the significant improvement achieved over three leading databases, and identify where each database is strongest in relation to XPath query performance.