Basic information about SDMX¶
What is SDMX?
SDMX (Statisical Data and Metadata eXchange) is a collaboration of some of the most important international organisations with the aim of encouraging and improving the exchange and sharing of statistical data and metadata.
SDMX is:
A Logical Model used to describe statistical data and also providing guidelines on how to structure the content.
An Architecture which enables efficient automated machine-to-machine exchange and sharing of data and metadata.
A technology that supports standardised information and methodological tools that can be used by all those involved in data exchange and processing.
Which benefits does it bring?
By accepting the use of a common description of the data, it is then used as a parameter guiding the exchange and processing of the data.
Data descriptions are made available to all, so that those interested in a certain topic can understand and use the data for different purposes.
SDMX is the leap from diverse and complex exchange systems to a common, harmonised and standardised exchange system.
The Standard
SDMX provides a way to model statistical data, structural metadata and the data exchange process and also defines a model for further explanatory metadata, the so-called reference metadata, which are generally in textual format.
In order to describe and document data, the standard refers to entities (in SDMX “artefacts”) that are organised in such a way as to represent and refer to the data appropriately:
The Id, agency and version are the three identifying elements of an SDMX artefact.
The Id is the identification code of the artefact.
Agency is the name of the organisation that is the creator and/or owner of the artefact.
Version gives the version of the artefact. In particular, if the artefact is finalised, it is not possible to modify it unless a new version of the artefact is created.
The main artefacts used to describe data and statistical metadata are:
Dataflow: structure describing the content of a dataset that the producing organisation provides for different reference time periods.
The characterising element of Dataflow, is the Data Structure Definition (DSD) as it defines the constitutive structure in terms of components (dimensions, attributes, measures).
NOTE:
Remember that the Dimensions are qualitative characteristics of the statistical units (e.g. reference period, reporting country, frequency, gender, …).
Attributes represent a qualitative characteristic of the observed data (confidentiality, status, version, decimals, unit of measure, table title, …), while Measures are the values of the observations.
Concept Scheme: is a grouping of concepts that refer to the components (dimensions, attributes, measures) of the DSD.
Code List: is a list of codes associated to components (dimensions and coded attributes) of the DSD.
Category Scheme: is a set of categories, hierarchically organised, that classify Dataflows.
These and other artefacts will be the focus of the ‘Data & Meta Manager’ tool.