3.8 Compatibility with Other Metadata Standards
The metadata standards outlined in this section are not utilized by the Bank for development data documentation (with the exception of SDMX for indicators’ structural metadata). Nonetheless, these standards are related and contribute to the broader metadata ecosystem for development data. Where appropriate and applicable, the Bank aligns its implementation of the metadata standards defined by the Chief Statistician with these other standards to maintain consistency.
3.8.1 Statistical Data and Metadata Exchange (SDMX)
The Statistical Data and Metadata eXchange (SDMX) [1], published as an ISO International Standard (ISO 17369), establishes a framework for exchanging statistical data and metadata using modern information technology. As one of the sponsoring organizations of SDMX, the World Bank is implementing SDMX for selected indicator datasets, with a strategic goal of ensuring SDMX compliance in the dissemination of all its time series data.
SDMX is primarily designed for machine-to-machine exchange of statistical data and metadata. Its core artefacts include:
Data Structure Definitions (DSDs): templates that specify the dimensions, attributes, and measures for a dataset (i.e., the structural metadata required for data exchange).
Metadata Structure Definitions (MSDs): templates that define the structure of metadata reports which contain descriptive and reference metadata.
Concepts and Code Lists: Concepts (e.g., “reference area,” “time period,” “unit,” “obs value”) are reusable semantic definitions that give meaning to dataset components; they are referenced in DSDs and MSDs. Code lists provide the permissible coded values for those concepts.
SDMX allows users to define MSDs as needed but does not dictate their exact content. To support thorough and useful descriptive and reference metadata for indicators, the Bank’s metadata standard offers a well-organized collection of suggested elements. As a result, SDMX and the Bank’s metadata standard work together in a complementary way, rather than duplicating efforts.
To facilitate SDMX integration, the Bank’s metadata standard and the Metadata Editor application have been designed with interoperability in mind:
The “data structure definition” section of the Bank’s indicator metadata standard allows for the export of SDMX-compliant DSDs.
The templates and controlled vocabularies developed in the Metadata Editor can be exported as SDMX-compliant MSDs.
Metadata generated using the Metadata Editor can be exported as SDMX-compliant metadatasets and reports.
3.8.2 DCAT, Schema.org, and Croissant
To improve how data can be discovered, shared, and accessed online, the Bank aligns its development data documentation with three complementary, widely adopted metadata standards: DCAT, Schema.org, and Croissant. While these are not used for managing metadata internally, they enhance interoperability with external applications and increase visibility through internet search engines and data catalogs. Ensuring compatibility with these generic standards keeps the Bank’s metadata ecosystem aligned with global data networks, making it easier for external users to find, access, and integrate Bank data across web platforms and search tools.”
DCAT (Data Catalog Vocabulary)[2] is a widely used vocabulary for publishing data catalogs on the Web, originally developed for government open data portals.
Schema.org[3] is a community-driven initiative founded by Google, Microsoft, Yahoo, and Yandex to develop a collection of schemas for structured data of any type published on the Internet. These schemas cover descriptive and metadata intended to improve search engine discoverability and facilitate metadata use and exchange.
Croissant[4] is a metadata format for machine learning datasets, developed as part of the MLCommons initiative by a collaboration of industry and academic partners. It extends Schema.org and provides a standardized, machine-readable structural metadata format for machine-learning (ML) datasets that makes datasets easier to discover, load across ML frameworks, and reuse with consistency.
The generation of metadata compliant with these additional, generic standards is automated. A defined subset of fields from the Bank’s specialized metadata standards is programmatically extracted and mapped to DCAT, Schema.org, and Croissant profiles as appropriate, producing compliant metadata artifacts without manual intervention by data curators. This approach minimizes burden on teams, ensures consistency, and enables continuous synchronization as source metadata evolves. Since DCAT, Schema.org, and Croissant metadata include only part of the canonical metadata, it is essential to maintain a clear connection to the definitive metadata records.
3.8.3 Open Archives Initiative - Protocol for Metadata Harvesting (OAI-PMH)
The Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) is a simple mechanism for repository interoperability. Data Providers are repositories that expose structured metadata via OAI-PMH. Service Providers then make OAI-PMH service requests to harvest that metadata. OAI-PMH is a set of six verbs or services that are invoked within HTTP.[5] The Bank seeks to develop data catalogs that comply with the OAI protocol. The metadata standards used by the Bank provide all necessary elements for compliance with OAI-PMH.