Key issues and priorities
The following topics have been identified so far by members of this committee:
- Establish "Practitioner Communities' for GSIM (and possibly GSBPM) to (eg) enhance user guides and maintain GSIM
- Evolve GSBPM to support CSPA
- Further define how DDI and SDMX can be used in a consistent manner to implement GSIM and ensure the interoperability of information produced by, and used by, different applications/services/repositories. Includes input to model-based DDI
- Defining profiles of DDI for more objects
- Determine, and then develop any specific standards required to support big data (includes explanation of the value of semantic technologies in interpreting big data)
- Further define the relationship between statistical community standards (eg GSIM, SDMX, DDI) and broader Linked Data / Semantic Web standards
- In particular, we think that that one of the goals to pursue consists in organizing metadata in SDMX according to the statistical role of metadata, as described in GSIM, creating appropriate concepts and code lists, and defining general DSDs structures. This step will help in connecting SDMX and DDI through GSIM.
- Another objective of modernisation could be to further develop the overarching items of GSBPM, namely quality and metadata, maybe in an annex by describing the main items in relation to the process phases
- The committee on standards should work efficiently with the other committees, especially with the Production and Methods Committee which has also a task on "common statistical production architecture". In order not to be overlaps and cross cutting subjects between them, some working procedures should be defined clearly (exp: the discussions in the METIS and MSIS meetings)
- To broaden the standardization concept to the NSOs, it can be defined in a clear manner what kind of activities should be done by NSOs for standardization (there are too many case studies in METIS, but more than this is needed)
- DDI is Ok for production process. Development of DDI also for analysis and dissemination processes to have one unique metadata repository can also be discussed.