An Efficient Storage for XML Data using  Relative Region  Coordinate

Dao Dinh Kha (Student No: 9951134)

Database and Media Systems Lab.


Master Thesis Abstract:

Since XML was invented, many index structures for XML data have been proposed. For these structures, update is a
problem because XML element's coordinates are expressed by absolute values. Due to the structural relationship among
elements in XML documents, we have to re-compute these absolute values if the content of source data is updated.
The reconstruction causes much time consumption, especially when XML data content is frequently updated. In this work
we propose an experimental indexing structure scheme based on the Relative Region Coordinate, that can effectively deal
with the content update problem. The main idea is that we express the coordinate of an XML element based on the region
of its parent element and store related coordinates together in a tree structure. In consequence, we need to update a
portion of index file in case of content updating. We also describe the outline of how to apply RRC method to an XML
indexing structure using Relational Database.