XML files are effective for data store, search and query when used as a single entity. However, when it comes search and query to multiple XML files, performance and efficiency are the ones that are to be degradated. To cope with such degradation, we are proposing an efficient way of merging XML files into a relational database for permanent storage and take advantage of relational database management system?s search and query features to locate certain record. Once the resulting record - used to be an XML file - is located, it will be converted back to an XML file for further use. Upon completion on further use, the XML is merged back to the database. To accommodate such need, multiple tables are created according to the XML file?s tag structure. In our approach, there are two data transformation functions -- one is to conver XML to relational data model and the other is relational data model to XML. We have developed a prototype conversion system for the technical feasibility analysis. For the sake of our experiment, we have used a virtual collaboration system called VCEI where for each discussion session, all the discussion or conference content are stored in a single XML file. Over the times, the number of discussions grows big and so is the number of XML files. The prototype also demonstrates the functional extension of the existing remote conferencing system.