Overview

To this point, the focus has been on applying modeling and architecture principles to XML prototypes and more importantly their representative W3C XML Schemas. In addition to these techniques, a process for schema design and engineering is needed. Given the rapid adoption of XML by the development community, the design and engineering of W3C XML Schemas has to date been somewhat ad hoc. When the decision to use XML has been made, the resulting structures and constraining schemas are often generated from an object-oriented model such as a class diagram. However, the development community may not have the expertise or the time to incorporate rigorous metadata characteristics or to apply critical data standards. Also, the responsibility for schema design and engineering may not have been formally defined, leading to a lack of rigorous metadata and related architectural practices.

The fundamental definition and capabilities of XML need to be considered with any design and engineering process. To repeat, XML is a descriptive metadata language and W3C XML Schemas provide a method of constraining XML content according to structure, organization, and metadata rules. What should be obvious to any practitioner is that effective engineering of a schema requires expertise with metadata and related data architecture disciplines. Without a metadata and data architecture focus, the broad proliferation of nonstandard and ad hoc XML transactions and schemas will be seen, resulting in increased data disparity, integration complexity, and a failure to reduce related technology costs.

Alternatively, the data architect must also recognize that...