| One of the challenges in managing XML data is to store and query the data efficiently. The query operations navigate XML data via path expressions and may lead to exhaustive searches because of little or no knowledge about the data in advance. Many index structures have been developed to address the problem. However, these indexing schemes either support a limited class of queries or can only be applied to some query processing stages. To overcome these limitations, the ToX project at the University of Toronto has developed ToXin, an index scheme that supports general path queries in all query processing stages. In this thesis we propose a scheme to map ToXin into relational databases and extend its current in-memory version to a persistent index. |
