Daniel W. Gillman
XML, Schemas, Names, and Confusion

gillman_d@bls.gov

This paper describes some general problems that are going to arise with the widespread use of XML. Many organizations and groups of organizations are developing XML Schemas for describing, managing, disseminating, and transferring data. Some of these schemas overlap with others either with by concept or by name. The namespace mechanism in XML disambiguates common names, but does not help in distinguishing the concepts that may be common across schemas. The problem is especially difficult when the names of similar elements are different.

XML registries, such as the OASIS/ebXML registry effort, allow developers to register schemas and instances, but not individual elements. A registry of elements is necessary to manage and determine the similarities and differences between elements in different schemas. Luckily, the ISO/IEC 11179 metadata registry standard is positioned to solve this problem. However, the standard by itself cannot do it. Detailed terminological systems (e.g., ontologies) are required to represent the relationships between concepts that are used in different schemas.

The paper will focus on describing these problems in more detail and describe the solutions that are requires as well.

Daniel W. Gillman
US Bureau of Labor Statistics
Phone +1.202.691.7523
FAX +1.202.691.7426