There is evidently a significant convergence between two important technologies: database technology and markup for the World Wide Web--in particular XML, the Extensible Markup Language (see http://www.w3.org/TR/REC-xml for the formal definition and Flynn (2000) for introductory material).
The Internet is just beginning to make an impact on the database world, and the fact that data as much as if not more so than documents will use XML to travel the Internet is just beginning to significantly influence the design and development of the XML family of standards.
Data means not just data in standard database formats (i.e. accessible via SQL) but also data in record structures and object hierarchies in application programs, each with idiosyncratic dump formats for archiving and interchange. The value of an XML representation for data of both kinds, both internally (for archiving and interchange) and externally (if you can show the world your data in the form of XML, they'll buy it) is becoming apparent in many commercial sectors (e-commerce not the least). The problem the work proposed here seeks to address is that a myriad of more or less ad-hoc mechanisms are springing up to manage the translation from internal data representations to XML and back again. We propose to use emerging international standards, in particular the W3C's XSLT and XML Schema (Clark 1999, Thompson et al. 2000), to provide a uniform, standards-based solution1.
The proposed work starts with the most tractable aspects of this problem to prototype a solution architecture, and progresses to address a number of more complex aspects, bringing to bear a number of strands of existing research and development work.