Bilkent University
Department of Computer Engineering
S E M I N A R

 

Log-based XML Storage Using Graph Partitioning

 

Reha Oğuz Selvitopi
MSc. Student
Computer Engineering Department
Bilkent University

Querying large XML documents efficiently requires constructing an index that covers whole XML data tree. The problem with constructing an index for large XML documents is that sometimes the index size can be very large so that it does not fit into memory. In such cases, XML data tree or its index must be partitioned and clustered across the disk pages. In our work, we partition XML data tree using graph partitioning with given workload and traversal information of past XPath queries and compare our method with XS algorithm (an algorithm for partitioning trees augmented with sibling edges) to minimize page faults for future XPath queries on XML data tree.

 

DATE: 9 November, 2009, Monday @ 16:20
PLACE: EA 409