Keywords: statistics, metadata, model-driven design, interoperability, schema design, standards, meta-model, Statistical Data and Metadata Exchange, SDMX, SDMX-EDI, SDMX-ML, time series, cross-sectional data, metadata, Web Services
Biography
Arofan Gregory is a principal at Aeon Consulting, and has been active in the Statistical Data and Metadata Exchange (SDMX) Case Study Project. Prior to that he was the product manager of the XML Common Business Library at Commerce One, and has been active in the ebXML Core Components work and earlier in the CommerceNet eCo initiative. Within the UBL initiative, he is an active member of the Naming and Design Rules subcommittee, the Context Methodology subcommittee, and the Library Content subcommittee; is the editor of the Context Classifications subcommittee, and is the chair of the Tools and Techniques subcommittee. He continues to be active in UN/CEFACT, working on the editorial team in the Core Components group. He has presented at various SGML and XML conferences, and has been active in this technology arena for the past decade, working in commercial publishing and as a consultant on document management and e-commerce applications.
The SDMX initiative - a joint effort by seven international organizations (OECD, BIS, ECB, Eurostat, World Bank, IMF, and the UN) - has just released version 1.0 of the SDMX standards. These standards describe W3C XML Schema and UN/EDIFACT formats (SDMX-ML and SDMX-EDI)for exchange of aggregated statistical data and related structural metadata. They also provide a meta-model for describing aggregated statistical data from any domain.
SDMX is a case study in using meta-model design to support interoperability and sophisticated code generation. The formats are all driven from a dedicated structural description based on a common meta-model. The model is expressed as a processable XML instance or EDI message. As a result, dedicated XML schemas, transformations, and other types of application-specific code can be automatically generated for specific data sets and to meet specific use-cases. Agreement at the level of the meta-model allows interoperability across this family of related format expressions. SDMX is also designed to be used with web-services technology, and includes guidelines for use with them.
From a business perspective, SDMX offers governmental and other organizations a standard for modelling and exchanging aggregated statistical data which is not domain-specific, and which supports the use of existing metadata vocabularies for statistical concepts. Formats are primarily designed around time-series views of data, but cross-sectional views are also supported. Several large-scale implementations are already planned.
In coming versions, efforts will be made to support descriptive metadata modelling and exchange, the use of a services-oriented architecture based on web services and registries, and more complete support for cross-sectional data and hypercubes. SDMX will also be working to support the harmonization of statistical concepts in future, as it relates to statistical data and metadata exchange.
Since this was a late-breaking talk, the author did not have time to complete the paper for the proceedings.
XHTML rendition made possible by SchemaSoft's Document Interpreter™ technology.