Conclusion

The fundamental characteristic of a native XML database is its pervasive use of the XML data model not only at the query language interface to the system, but in all its other interfaces as well. We have reviewed some of these interfaces and seen how XQuery represents only a small part of the facilities that a real system needs to offer.

Because XML is being stored and manipulated in databases, a new set of database features attuned to XML is expected in a native XML DBMS. This is partly because the data is semi-structured—as well as structured data, it may contain loosely structured document-centric information that requires special features not found in relational or object databases. Even data-centric XML, containing information such as invoices and purchase orders, is very different from relational data because of the tree-based XML data model.

With data definition, update capability, and full-text search added to XQuery, along with other XML technologies integrated into a native XML database, such a DBMS becomes a natural choice for managing XML data, since it can so easily work with XML data. Such systems will initially be used in areas where relational databases are weak, that is, where the data structures are at the same time complex, variable, and flexible. In time, with a flexible data model like XML and a powerful language like XQuery coupled with other capabilities such as searching technologies, native XML database management systems may become useful in areas that are currently the exclusive preserve of the relational database.