Mapping Database Content to XML Pages - A Metadata-Based Approach

Today's web-based information systems usually employ database systems to store their content, forming the basis for web page construction. Generating web pages on request, especially in peak times, however, can lead to severe performance problems. Thus, our approach of "self maintaining" web pages suggests to materialize web pages outside the database and to update them every time the underlying content changes. Maintaining the mapping knowledge between database content and web pages in an efficient way, however, represents a major challenge. Our approach of self-maintaining web pages tackles this problem by means of different concepts. First, instead of allocating individual content fragments to single web pages, parameterized content fragments are allocated to web page classes. Second, the Extensible Markup Language (XML) is used as an intermediate layer between the database content and the final web pages, thus allowing to propagate updates only to those elements which are actually affected. Third, a meta schema allows to reify the mapping knowledge, thus enhancing maintainability and changeability. This paper focuses on the meta schema, representing a core component of our approach.