A Hybrid Method of Storing XML Data Using RDBMS

As the Web-based e-Business prevails, the volume of XML data on the Web is getting larger than ever. Although much research has been done on decomposing and storing XML data in RDB, which is now the most popular storage for XML, and on processing XML queries through their SQL counterparts, little attention was paid to how to alleviate the burden of storing massive volume of XML data. In this paper, we propose a hybrid method of storing XML data in RDB, whereby the unit of storage could be an XML subtree as well as an XML node. The part of XML data whose nodes were separately stored could be reformed into an XML subtree for storing when it gets rarely queried or less valuable for reference as time goes by. With this method, we designed and implemented a hybrid XML storage and query processing system, comparing it with the conventional system where an XML node is the only unit of storage. Through experiments, we compared storage efficiency and query processing performance, validating the effectiveness of our proposed system.