Linked SDMX Data

Sarven Capadisli1
Sören Auer2
Axel-Cyrille Ngonga Ngomo3
  • 1Leipzig_University, Institut für Informatik, AKSW, Postfach 100920, D-04009 Leipzig, Germany
  • 2Leipzig_University, Institut für Informatik, AKSW, Postfach 100920, D-04009 Leipzig, Germany
  • 3Leipzig_University, Institut für Informatik, AKSW, Postfach 100920, D-04009 Leipzig, Germany
CC BY 4.0
In Reply To
Semantic Web Journal Call: 2nd Special Issue on Linked Dataset Descriptions
Appeared In
SWJ-LDD-2015 (Semantic Web, Linked Dataset Descriptions 2015) , Tracking 454-1631 , DOI 10.3233/SW-130123 , ISSN 2210-4968 , Volume 6 , Issue 2 , Pages 105112
Path to high fidelity Statistical Linked Data


As statistical data is inherently highly structured and comes with rich metadata (in form of code lists, data cubes etc.), it would be a missed opportunity to not tap into it from the Linked Data angle. At the time of this writing, there exists no simple way to transform statistical data into Linked Data since the raw data comes in different shapes and forms. Given that SDMX (Statistical Data and Metadata eXchange) is arguably the most widely used standard for statistical data exchange, a great amount of statistical data about our societies is yet to be discoverable and identifiable in a uniform way. In this article, we present the design and implementation of SDMX-ML to RDF/XML XSL transformations, as well as the publication of OECD, BFS, FAO, ECB, IMF, UIS, FRB, BIS, and ABS dataspaces with that tooling.

Categories and Subject Descriptors

  • H.4 [Information Systems Applications]: Linked Data
  • D.2 [Software Engineering]: Semantic Web



While access to statistical data in the public sector has increased in recent years, a range of technical challenges makes it difficult for data consumers to tap into this data at ease. These are particularly related to the following two areas:

  • Automation of data transformation of data from high profile statistical organizations.
  • Minimization of third-party interpretation of the source data and metadata and lossless transformations.

Development teams often face low-level repetitive data management tasks to deal with someone else's data. Within the context of Linked Data, one aspect is to transform this raw statistical data (e.g., SDMX-ML) into an RDF representation in order to be able to start tapping into what's out there in a uniform way.

The contributions of this article are two-fold. We present an approach for transforming SDMX-ML based on XSLT 2.0 templates and showcase our implementation which transforms SDMX-ML data to RDF/XML. Following this, SDMX-ML data from OECD (Organisation for Economic Co-operation and Development), BFS (Bundesamt für Statistik@de, Swiss Federal Statistical Office@en), FAO (Food and Agriculture Organization of the United Nations), ECB (European Central Bank), and IMF (International Monetary Fund), UIS (UNESCO Institute for Statistics), FRB (Federal Reserve Board), BIS (Bank for International Settlements), ABS (Australian Bureau of Statistics) are retrieved, transformed and published as Linked Data.


As pointed out in Statistical Linked Dataspaces (Capadisli, S., 2012), what linked statistics provide, and in fact enable, are queries across datasets: Given that the dimension concepts are interlinked, one can learn from a certain observation's dimension value, and enable the automation of cross-dataset queries.

Moreover, a number of approaches have been undertaken in the past to go from raw statistical data from the publisher to linked statistical data, as discussed in great detail in Official statistics and the Practice of Data Fidelity (Cyganiak, R., 2011). These approaches go from retrieval of the data by majority; in tabular formats: Microsoft Excel or CSV, tree formats: XML with a custom schema, SDMX-ML, PC-Axis, to transformation into different RDF serialization formats. As far as graph formats go, majority of datasets in those formats not published by the owners. However, there are number of statistical linked dataspaces in the LOD Cloud already.

A number of transformation efforts are performed by the Linked Data community based on various formats. For example, the World Bank Linked Dataspace is based on custom XML that the World Bank provides through their APIs with the application of XSL Templates. The Transparency International Linked Dataspace's data is based on CSV files with the transformation step through Google Refine and the RDF Extension. That is, data sources provide different data formats for the public, with or without accompanying metadata e.g., vocabularies, provenance. Hence, this repetitive work is no exception to Linked Data teams as they have to constantly be involved either by way of hand-held transformation efforts, or in best-case scenarios, it is done semi-automatically. Currently, there is no automation of the transformation step to the best of our knowledge. This is generally due to the difficulty of the task when dealing with the quality and consistency of the statistical data that is published on the Web, as well as the data formats that are typically focused on consumption. Although SDMX-ML is the primary format of the high profile statistical data organizations, it is yet to be taken advantage of.

SDMX-ML to Linked Data

Recently, SDMX is approved by ISO as International Standard. It is a standard which provides the possibility to consistently carry out data flows between publishers and consumers. SDMX-ML (using XML syntax) is considered to be the gold standard for expressing statistical data. It has a highly structured mechanism to represent statistical observations, classifications, and data structures. Organizations behind SDMX are BIS (Bank for International Settlements), OECD, UN (United Nations), ECB, World Bank, IMF, FAO, and Eurostat.

We argue that high-fidelity statistical data representation in Linked Data should take advantage of SDMX-ML as it is widely adopted by data producers with rich data about our societies, making the need for transforming SDMX-ML to RDF and publishing accompanying Linked Dataspaces of paramount importance

Data Sources

As a demonstration of the SDMX-ML to RDF transformations, the selection of datasets here are from the following organizations. Instead of regurgitating publisher's profile about themselves, and to keep this part brief, I'll quote the first sentences from their about pages:

The mission of the Organisation for Economic Co-operation and Development (OECD) is to promote policies that will improve the economic and social well-being of people around the world from OECD Our mission.
Swiss Statistics, the Federal Statistical Office’s web portal. Offering a modern and appealing interface, our website proposes a wide range of statistical information on the most important areas of life: population, health, economy, employment, education and much more from BFS Welcome.
Achieving food security for all is at the heart of FAO's efforts - to make sure people have regular access to enough high-quality food to lead active, healthy lives from FAO's mandate.
Whose main task is to maintain the euro's purchasing power and thus price stability in the euro area from ECB home.
Working to foster global monetary cooperation, secure financial stability, facilitate international trade, promote high employment and sustainable economic growth, and reduce poverty around the world from IMF home.
The primary source for cross-nationally comparable statistics on education, science and technology, culture, and communication for more than 200 countries and territories from UIS home.
The Federal Reserve, the central bank of the United States, provides the nation with a safe, flexible, and stable monetary and financial system from FRB home.
The Bank for International Settlements (BIS) is an international organisation which fosters international monetary and financial cooperation and serves as a bank for central banks from BIS home.
We assist and encourage informed decision making, research and discussion within governments and the community, by leading a high quality, objective and responsive national statistical service from ABS mission statement.

The OECD, FAO, ECB, IMF, UIS, BIS, ABS datasets consisted of observational and structural data. The OECD, ECB, UIS, BIS, ABS provided complete coverage (to the best of our knowledge), whereas FAO had partial fishery related data, and IMF partial data over their REST service. BFS had all of their classifications available, with no observational data in SDMX-ML.

Data Retrieval

As SDMX-ML publishers have their own publishing processes, availability and accessibility of the data varied. After obtaining the dataset codes, names, and URLs with common Linux command-line work, a Bash script was created to retrieve the data.

On how to retrieve all of the datasets from the OECD website was not entirely clear. In order to automatically get a hold of list of datasets, I copied the innerHTML of the DOM tree that contained all the dataset codes from OECD.StatExtracts to a temporary file. This was done due to the fact that a simple scrape of an HTTP GET wasn't possible as the data on the page was populated via JavaScript on document ready. After constructing a list of datasets and structures to get, two REST API endpoints were called.

BFS offered a Microsoft Excel document which contained a catalog of their classifications and URLs for retrieval.

After searching for keywords along the lines of SDMX at a search-engine nearby, FAO Fisheries and SDMX Registry and Repository, and its children pages were marked for SDMX-ML retrieval.

ECB had a similar REST API to OECD. Additionally, SDMX Dataflows was retrieved to get a primary list of datasets to retrieve. Some of the large datasets was retrieved by making multiple smaller calls to the API using a call per refereance area.

IMF had the same procedure as ECB and OECD.

FRB had the same procedure as above.

UIS had the same procedure as above.

BIS had the same procedure as above.

ABS had the same procedure as above.


We now discuss provenance at: retrieval, transformation, and post-processing.

Provenance at Retrieval

At the time of data retrieval, information pertaining to provenance was captured using the PROV Ontology in order to further enrich the data. This RDF/XML document contains prov:Activity information which indicates the location of the XML document on the local filesystem. It contains other provenance data like when it was retrieved, with what tools, etc. This provenance data from retrieval may be provided to the XSL Transformer during the transformation phase and VoID enrichment.

Provenance at Transformation

Resources of type qb:DataStructureDefinition, qb:DataSet, skos:ConceptScheme are also typed with the prov:Entity class. Also properties prov:wasAttributedTo were added to these resources with the creator value which is of type prov:Agent obtained from XSLT configuration. There is a unique prov:Activity for each transformation, and it has a schema:name, and contains values for prov:startedAtTime, prov:wasAssociatedWith (the creator), prov:used (i.e., source XML, XSL to transform) to what was prov:generated (and source data URI that it prov:wasDerivedFrom). It also declares dcterms:license where value taken from XSLT configuration. The provenance document from the retrieval phase may be provided to the transformer. In this case, it establishes a link between the current provenance activity (i.e., the transformation), with the earlier provenance activity (i.e., the retrieval) using the prov:wasInformedBy property.

Provenance at Post-processing

The post-processing step for provenance is intended to retain provenance data for future use. As datasets get updated, it is important to preserve information about past activities by way of exporting all instances of the prov:Activity class from the RDF store. Activities are unique artifacts, on a conceptual level as well as with regard to referencing them. Since one of the main concerns of provenance is to keep track of activities, this post-processing step also allows us to retain a historical account of all activities during the data lifecycle, and to preserve all previously published URIs (cf. Cool URIs don't change).

Data Preprocessing

By in large, there was no need to pre-process the data as the transformation dealt with the data as it was. However, some non-vital SDMX components were omitted from the output. For instance, one type of attribute in OECD and ECB observations contained free-text as opposed to its corresponding code from a codelist. Since the RDF Data Cube required codes as opposed to free-text for dimension values, some attributes were excluded. The decision here was to trade-off some precision in favour of retaining the dataset.

Data Modeling

This section goves over several areas which are at the heart of representing statistical data in SDMX-ML as Linked Data. The approach taken was to provide a level of consistency for data consumers and tool builders for all statistical Linked Data with its origins from data in SDMX-ML.


Besides the common vocabularies: RDF RDFS, XSD, OWL, XSD, the RDF Data Cube vocabulary is used to describe multi-dimensional statistical data, and SDMX-RDF for the statistical information model. PROV-O is used for provenance coverage. SKOS and XKOS to cover concepts, concept schemes and their relationships to one another.


SDMX data publishers version their classifications and the generated cubes refer to particular versions of those classifications. Consequently, versions need to be explicitly part of classification URIs in order to uniquely identify them. Although including version information in the URI is disputed by some authors, it is a good exception for identifying different concepts and data structures. Jeni Tennison et al discussed Versioning URIs, and concluded that there was no one-size-fits all solution. An alternative approach using named graphs for a series of changes was proposed in Linking UK Government Data.

URI Patterns

An outline for the URI patterns is given in table below: authority is replaced with the domain (see also: Agency identifiers and URIs) followed with class, code, concept, dataset, property, provenance, or slice as example. These tokens as well as / which is used to separate the dimension concepts in URIs can be configured in the toolkit.

In order to construct the URIs for the above patterns, some of the data values are normalized to make them URI safe but not altered in other ways (e.g., lower-casing). The rationale for this was to keep the consistency of terms in SDMX and RDF.

URI Patterns
Entity typeURI Pattern


XSD datatypes are assigned to literals are based on the value of the measure component (e.g., decimal, year). In the absence of this datatype, observation values are checked whether they can be casted to xsd:decimal. Otherwise, they are left as plain literals.

Linked SDMX Data Transformation

The Linked SDMX XSLT 2.0 templates and scripts are developed to transform SDMX-ML data and metadata to RDF/XML. Its goals are:

  • To improve access and discovery of cross-domain statistical data.
  • To perform the transformation in a lossless and semantics preserving way.
  • To support and encourage statistical agencies to publish their data using RDF and integrating the transformation into their workflow.

The key advantage of this transformation approach is that additional interpretations are not required by the data modeler especially in comparison to alternative transformation (e.g., CSV or XML to RDF serialization). Since the SDMX-RDF vocabulary is based on SDMX-ML standard, and the RDF Data Cube vocabulary is closely aligned with the SDMX information model, the transformation is to a large extent a matter of mapping the source SDMX-ML data to its counter parts in RDF.

Features of the transformation

  • Transforms SDMX KeyFamilies, ConceptSchemes and Concepts, CodeLists and Codes, Hierarchical CodeLists, DataSets.
  • Configurability for SDMX publisher's needs.
  • Detection and referencing CodeLists and Codes of external agencies.
  • Support of interlinking publisher-specific annotation types.
  • Support for omission of components.
  • Inclusion of provenance data.
Transformation Process

What is inside?

It comes with scripts and sample data:

  • XSLT 2.0 templates to transform Generic SDMX-ML data and metadata. It includes the main XSL template for generic SDMX-ML, an XSL for common templates and functions, and an RDF/XML configuration file to set preferences like base URIs, delimiters in URIs, how to map annotation types.
  • Bash script that transforms sample data using saxonb-xslt.
  • Sample SDMX Message and Structure retrieved from those organizations that are initially involved in the SDMX standard, as well as from BFS.


The requirements for the Linked SDMX toolkit are an XSLT 2.0 processor to transform, and optionally to configure some of the settings in the transformation with provided config.ttl (in RDF Turtle) and transforming that to an abbreviated version of RDF/XML. In sequel some of they key features are described in more detail.

The transformation follows some common Linked Data practices as well as other ones out of thin air.


The config file is used to pre-configure some of the decisions that are made in XSL templates. Here is an outline for some of the noteworthy things.

Agency identifiers and URIs

agencies.ttl is used to track some of the mappings for maintenance agencies. It includes the maintenance agency's i.e., the SDMX publisher's, identifier that's in the SDMX Registry, as well as the base URI for that agency. This file allows references to external agency identifiers to be looked up for their base URI and used in the transformations. Currently this agency recognition is treated as either "SDMX" or some agency that's publishing the actual statistics.

In the case of SDMX, when there is a reference to SDMX CodeLists and Codes, it is typically indicated by the component agency being set to SDMX e.g., codelistAgency="SDMX" of a structure:Component and/or agencyID="SDMX" of a CodeList with id="CL_FREQ". When this is detected, corresponding URIs from the SDMX-RDF vocabulary is used e.g., for metadata;, and data;

Similarly, an agency might use some other agency's codes. By following the same URI pattern conventions, the agency file is used to find the corresponding base URI in order to make a reference. For example, here is a coded property that's used by European Central Bank (ECB) to associate a code list that's defined by Eurostat (eurostat):


Naturally, the transformation does not re-define metadata that's from an external agency as the owners of the data would define them under their authority.

URI configurations

Base URIs can be set for classes, codelists, concept schemes, datasets, slices, properties, provenance, as well as for the source SDMX data.

The value for uriThingSeparator e.g., /, lets one set the delimiter to separate the "thing" from the rest of the URI. This is typically either a / or #. For example, if slash is used, an URI would end up like http://{authority}/code/{version}/CL_GEO (note the last slash before CL_GEO). If hash is used, an URI would end up like http://{authority}/code/{version}#CL_GEO.

Similarly, uriDimensionSeparator can be set to separate dimension values that's used in RDF Data Cube observation URIs. As observation should have its own unique URI, the method to construct URIs is done by taking dimension values as safe terms to be used in URIs separated by the value in uriDimensionSeparator. For example, here is a crazy looking observation URI where uriDimensionSeparator is set to /: http://{authority}/dataset/HEALTH_STAT/EVIEFE00/EVIDUREV/AUS/1960. But with uriThingSeparator set to # and uriDimensionSeparator set to -, it could end up like http://{authority}/dataset/HEALTH_STAT#EVIEFE00-EVIDUREV-AUS-1960. HEALTH_STAT is the dataset id.

Creator's URI can also be set which is also used for provenance data.

Default language

From the configuration, it is possible to force a default xml:lang on skos:prefLabel and skos:definition when lang is not originally in the data. If config.rdfcontains a non-empty lang value it will use it. Default language may also be applied in the case of Annotations. See Interlinking SDMX Annotations for example.

Interlinking SDMX Annotations

SDMX Annotations contain important information that can be put to use by the publisher. Data in AnnotationTypes are typically used as publisher's internal conventions. Hence, there is no standardization on how they are used across all SDMX publishers. In order not to leave this information behind in the final transformation, the configuration allows publishers to define the way they should be transformed. This is done by setting interlinkAnnotationTypes: the AnnotationType to detect (in rdfs:type), the predicate (as an XML QName) to use (in rdf:predicate), whether to apply instances of Concepts or Codes to apply to, or as Literals (in rdf:range), and whether to target AnnotationText or AnnotationTitle (in rdfs:label). Currently this feature is only applied to Annotations in Concepts and Codes. Only the AnnotationTypes with a corresponding configuration will be applied, and unspecific ones will be skipped.

Omitting components

There are cases in which certain data parts contain errors. To get around this until the data is fixed at source, and without giving up on rest of the data at hand, as well as without making any significant assumptions or changes to the remaining data, omitComponents is a configuration option to explicitly skip over those parts. For example, if the Attribute values in a DataSet don't correspond to coded values - where they may contain whitespace - they can be skipped without damaging the rest of the data. This obviously gives up on precision in favour of still making use of the data.

Linked Datasets

This section provides information on the publication of OECD, BFS, FAO, ECB, UIS, FRB, and BIS datasets.

The original SDMX-ML files were transformed to RDF/XML using XSLT 2.0. Saxon’s command-line XSLT and XQuery Processor tool was used for the transformations, and employed as part of Bash scripts to iterate through all the files in the datasets.

RDF Datasets

Here are some statistics for the transformations.

The command-line tool saxonb-xslt was used to conduct the XSL transformations. 12000M of memory was allocated on a machine with Intel(R) Xeon(R) CPU E5620 @ 2.40GHz. Linux kernel 3.2.0-33-generic was used. Table [Transformation time] provides information on datasets; input SDMX-ML size, output RDF/XML size, their size difference in ratio, and the total amount transformation time.

Transformation time
DatasetInput sizeOutput sizeRatioTime
Input size (rounded) refers to the original data in XML, and the output (rounded) is the RDF/XML size. Time is the real process time.
OECD3400 MB24000 MB1:7.1131m25.795s
BFS111 MB154 MB1:1.42m38.225s
FAO902 MB4400 MB1:4.931m48.207s
ECB11000 MB35000 MB1:3.2316m46.329s
IMF392 MB3400 MB1:8.728m11.826s
UIS115 MB896 MB1:7.82m33.214s
FRB783 MB11000 MB1:14.196m20.676s
BIS845 MB4000 MB1:4.721m7.203s
ABS34000 MB160000 MB1:4.7831m52.558s

Table [Transformed data] provides data on the transformed data; number of triples it contains, as well as the number of qb:Observations, and the ratio.

Transformed data
DataNumber of triplesNumber of observationsRatio
Metadata (from graph/meta) includes dataset structures and classifications. Ratio refers to rounded ratio of total number of triples (rounded) to number of observations (rounded) in the dataset.
OECD Dataset305 million30 million10.2:1
OECD Metadata1.15 millionN/AN/A
BFS Metadata1.5 millionN/AN/A
FAO Dataset53 million7.2 million7.4:1
FAO Metadata0.37 millionN/AN/A
ECB Dataset469 million26 million18:1
ECB Metadata0.47 millionN/AN/A
IMF Dataset36 million3.3 million10.9:1
IMF Metadata0.05 millionN/AN/A
UIS Dataset10.4 million1.4 million7.4:1
UIS Metadata0.09 millionN/AN/A
FRB Dataset135 million9.8 million13.8:1
FRB Metadata0.04 millionN/AN/A
BIS Dataset54 million3.6 million15:1
BIS Metadata0.04 millionN/AN/A
ABS Dataset2360 million160 million14.8:1
ABS Metadata6.06 millionN/AN/A

Table [Resource counts] provides further statistics on prominent resources. It gives a contrast between the classifications and the dataset.

Resource counts
Source datasetskos:ConceptSchemeskos:Conceptrdf:Propertyqb:Observation
Count of resources in datasets


Initial interlinking is done among the classifications themselves in the datasets. The OECD and UIS classifications in particular contained highly similar codes (in some cases the same) throughout its codelists. Majority of the interlinks are between skos:Concepts with link relation skos:exactMatch.

The consequent interlinking was done with DBpedia, World Bank Linked Data, Transparency International Linked Data, and EUNIS using LInk discovery framework for MEtric Spaces (LIMES): A Time-Efficient Hybrid Approach to Link Discovery (Ngonga Ngomo, A.-C., 2011)

Figure [SDMX concept links] gives an overview of the complete connectivity of a concept that's linked internally, externally, and with sdmx-codes where applicable, as well as the interlinking that was done to an external concept.

RDF Data Storage

Apache Jena’s TDB storage system is used to load the RDFized data using TDB’s incremental tdbloader utility. tdbstats, the tool for TDB Optimizer is executed after a complete load to internally update the count of resources in order for TDB to make the best decision to come up with future query results.

Individual datasets from each organization were transformed to N-Triples format before loading into the store. Each RDF Data Cube dataset was imported into its own NAMED GRAPH in the store. Given the significant load speed on an empty database, N-Triples were ordered from largest to smallest, and then loaded.


The publication steps are described in this section.

Dataset Discovery and Statistics

As VoID file is generally intended to give an overview of the dataset metadata i.e., what it contains, ways to access it or query it, each dataspace contains Vocabulary of Interlinked Datasets (VoID) files accessible through their .well-known/void. Each OECD, BFS, FAO, ECB, IMF, UIS, FRB, BIS, ABS VoID contains locations to RDF datadumps, named graphs that are used in the SPARQL endpoint, used vocabularies, size of the datasets, interlinks to external datasets, as well as the provenance data which was gathered through the retrieval and transformation process. The VoID files were generated automatically by first importing the LODStats information into respective graph/void into the store, and then a SPARQL CONSTRUCT query to include all triples as well as additional ones which could be actively created based on the available information in all graphs.

Dataset statistics are generated and are included in the VoID file using LODStats, LODStats – An Extensible Framework for High-performance Dataset Analytics (Demter, J., 2012).


The HTML pages are generated by the Linked Data Pages framework, where Moriarty, Paget, and ARC2 does the heavy lifting for it. Given the lessons learned over the years about Linked Data publishing, there is a consideration to either take Linked Data Pages further (originally written in 2010), or to adapt one of the existing frameworks after careful analysis.

SPARQL Endpoint

Apache Jena Fuseki is used to run the SPARQL server for the three datasets. SPARQL Endpoints are publicly accessible and read only at their respective /sparql and /query locations for OECD, BFS, FAO, ECB, IMF, UIS, FRB, BIS, ABS. Currently, 12000MB of memory is allocated for the single Fuseki server running all datasets.

Data Dumps

The data dumps for the datasets are available from their respective /data/ directories: OECD, BFS, FAO, ECB, IMF, FRB, UIS, BIS, ABS. Additionally, they are mentioned in the VoID files. The the Data Hub entries (see below) also contains links to the dumps.

Source Code

The code for transformations is at csarven/linked-sdmx, and for retrieval and data loading to RDF store for OECD is at csarven/oecd-linked-data, for BFS is at csarven/bfs-linked-data, for FAO is at csarven/fao-linked-data, for ECB is at csarven/ecb-linked-data, for IMF is at csarven/imf-linked-data, for UIS is at csarven/uis-linked-data, for FRB is at csarven/frb-linked-data, for ABS is at csarven/abs-linked-data. All using the Apache License 2.0.

Data License

All published Linked Data adheres to original data publisher’s data license and terms of use. Additionally attributions are given on the websites. The Linked Data version of the data is licensed under CC0 1.0 Universal (CC0 1.0) Public Domain Dedication.

Announcing the Datasets

For other ways for these datasets to be discovered, they are announced at mailing lists, status update services, and at the Data Hub: OECD is at oecd-linked-data, BFS is at bfs-linked-data, FAO is at fao-linked-data, ECB is at ecb-linked-data, IMF is at imf-linked-data, UIS is at uis-linked-data, FRB is at frb-linked-data, BIS is at bis-linked-data, ABS is at abs-linked-data.


With this work we provided an automated approach for transforming statistical SDMX-ML data to Linked Data in a single step. As a result, this effort helps to publish and consume large amounts of quality statistical Linked Data. Its goal is to shift focus from mundane development efforts to automating the generation of quality statistical data. Moreover, it facilitates to provide RDF serializations alongside the existing formats used by high profile statistical data owners. Our approach to employ XSLT transformations does not require changes to well established workflows at the statistical agencies.

One aspect of future work is to improve the SDMX-ML to RDF transformation quality and quantity. Regarding quality, we aim to test our transformation with further datasets to identify shortcomings and special cases being currently not yet covered by the implementation. Also, we plan the development of a coherent approach for (semi-)automatically interlinking different statistical dataspaces, which establishes links on all possible levels (e.g. classifications, observations). With regard to quantity, we plan to publish statistical dataspaces for Bank for International Settlements (BIS), World Bank and Eurostat based on SDMX-ML data.

The current transformation is mostly based on the generic SDMX format. Since some of the publishers make their data available in compact SDMX format, the transformation toolkit has to be extended. Alternatively, the compact format can be transformed to the generic format first (for which tools exist) and then Linked SDMX transformations can be applied. Ultimately, we hope that Linked Data publishing will become a direct part of the original data owners workflows and data publishing efforts. Therefore, further collaboration on this will expedite the provision of uniform access to statistical Linked Data.


We thank Richard Cyganiak for his ongoing support, as well as graciously offering to host the dataspaces on a server at Digital Enterprise Research Institute. We also acknowledge the support of Bern University of Applied Sciences for partially funding the transformation effort for the pilot Swiss Statistics Linked Data project and thank Swiss Federal Statistical Office for the excellent collaboration from the very beginning.


8 interactions

Benedikt’s photoBenedikt replied on

Nicely done, Sarven! A TOC at the beginning of the article would be nice.



Ana Ulu replied on

Thanks for this sharing. I also interest in SDMX. Hope could implement this to our statistics organization, especially in my country

Anonymous Reviewer replied on

This paper presents transformation from the ISO standard statistical data format SDMX-ML to RDF. URI patterns, interlinking, and publication is proposed for several statistical datasets. The paper is clearly written and the tools can be used for publishing other datasets, although some extensions or specific configuration might be necessary.

The only comments are that the paper is 2 pages too long according to the call. Also it would be better if datasets were published by the providers.

Anonymous Reviewer replied on

This paper describes a practical workflow for transforming SDMX collections into Linked Data. The authors focus on four relevant statistical datasets:

- OECD: whose mission is to promote policies that will improve the economic and social well-being of people around the world. - BFS Swiss Statistics, due to the Federal Statistical Office's web portal offering a wide range of statistical information including population, health, economy, employment and education. - FAO, which works on achieving food security for all to make sure people have regular access to enough high-quality food. - ECB, whose main task is to maintain the euro's purchasing power and thus price stability in the euro area.

Nevertheless, the tool proposed in the paper can be easily used for transforming any other SDMX dataset into Linked Data.

On the one hand, statistical data are rich sources of knowledge currently underexploited. Any new approach is welcome and it describes an effective workflow for transforming SDMX collections into Linked Data. On the other hand, the approach technically sounds. It describes a simple but effective solution based on well-known toolss, guaranteing robustnees and making easy its final integration into existing environments. Thus, the workflow is a contribution by itself and each stage describes how it impacts in the final dataset configuration.

With respect to the obtained datasets, these are clearly described in Sections 7 and 8. These reuse well-known vocabularies and provide interesting interlinkage between them but also with DBpedia, World-Bank, Transparency International and EUNIS. Apache Jena TDB is used to load RDF and Apache Jena Fuseki is used to run the SPARQL endpoint. Datasets are also released as RDF dumps (referenced from Data Hub).

Finally, its relevant for me how scalability problems are addressed, because I think that 12 GB is an excessive amount of memory for processing these datasets (the largest one outputs less 250 million triples). Do you have an alternative for processing larger datasets? Maybe you could partition the original dataset into some fragments: is the tool enough flexible to support it? Please, explain how scalability issues will be addressed to guarantee big datasets to be effectively transformed.

Anonymous Reviewer replied on

This paper describes the Linked Data datasets, and the process used to generate them, for a set of SDMX-enabled datasets coming from relevant international and country-focused statistical organisations (even some work has been done during the review process on another IMF dataset).

The paper is clearly fitting the special issue call, and the datasets are of good quality, and are made available under open licenses in such linked data format. I particularly like the extensive use of DataCube and Prov-o, and the approach taken for versioning of codelists and concept schemes.

The comments provided in this review are mostly curiosities about some design decisions or requests to make things more clear:

  • First of all, a large part of the paper is about the process used to generate the Linked data datasets, but not so much about the datasets themselves or the problems associated to problems in the original SDMX data, which certainly exist still in current sdmx implementations.
  • One example is related to the management of some NL descriptions of codes in codelists. These are simply ignored. Could they be handled differently? How manty of the datasets present this situation?
  • URI patterns. I find them nicely created and sensible. However, I would have some questions: how do you order dimensions in observations? are you using order in the dimensions in the data cube structures? Why in the owl:Class part do you only use codelistID instead of considering also conceptIDs?
  • The interlinkage of sdmx annotations part should be better explained, probably with an example that illustrates how it works.
  • The interlinking of datasets should be better explained as well. are all found links correct? I have found many cases of compatible codelists that are produced as separate codelists by stat offices and hence in principle are not linked, but could be. Have you found those cases?

Jay Devlin’s photoJay Devlin replied on

Great work Sarven, I'm keen to do more work on this. I'm working for Stats NZ at the moment but won't be for too long as I've moved to Ireland. But with the emergence of the SDMX Global Registry soon you should be able to collect a lot of interlinking data from that using the SDMX StructureSet structures which can hold both a ConceptMap and a CodeMap; so you could theoretically create your skos:exactMatch directly from these structures. Once the global registry is in place it obviously needs the key players to populate their data collaboratively.

We are building an RDF based classification system in SNZ which should be able to enrich the experience of doing this sort of thing in the future.

I'll have a play at this trying to get a few of our SDMX Datasets into the RDF and will be in touch. We have an SDMX Registry in v2.1 with the data exposed using v2.0 but I'd like to get the transforms working with v2.1 for at least the structure (and in preparation for the consolidated messages introduced in v2.1).

Cheers ...J