Share and reuse interoperability solutions for public administrations

The Asset Description Metadata Schema (ADMS) v1.00 is a metadata vocabulary to describe semantic standards (e.g. core components, domain models, schemas, thesauri, taxonomy, or code lists) on the Web. This article shows how you can generate ADMS descriptions of your semantic assets from a spreadsheet using the RDF extension to Google Refine  released by DERI‘s Fadi Maali and Richard Cyganiak.

On April 18, ADMS v1.00 was released on Joinup (see announcement by WG Chair), under the ISA Open Metadata Licence v1.1. ADMS is an initiative of the ISA Programme of the European Commission [3] which formed a multidisciplinary working group of researchers, EU institutions, standardisation bodies and other experts.

ADMS v1.00 helps you document the features of your semantic standard such as its name, description, theme, asset type, representation technique, origin, publisher in a machine-readable way, which can be either ADMS XML or ADMS RDF format. Such machine-readable metadata in ADMS format brings the following benefits:

• De-centralised publishing: ADMS allows you to describe your semantic asset once and use this description everywhere.
• Federation of repositories:  ADMS allows interoperability platforms, such as Joinup, to aggregate description metadata from standardisation organisations, national semantic asset repositories, etc. The promo video of the ADMS-enabled federation nicely visualises this.
• Cross-repository querying: ADMS is a common description format of semantic assets and will allow searching for semantic assets located on disparate repositories.
• Faceted search: ADMS is a rich metadata vocabulary which allows new user-interface techniques such as faceted search.

This article shows how you can generate ADMS descriptions of your semantic assets from a spreadsheet using the RDF extension to Google Refine. This can be done in four steps:

1. Use the spreadsheet template to describe your semantic standard;
2. Import your spreadsheet in Google Refine;
3. Add the pre-defined mappings using the Google Refine RDF extension; and
4. Export to RDF using the Google Refine RDF extension.

1. Use the spreadsheet template to describe your semantic standard

The spreadsheet template contains all concepts and fields in the ADMS v1.00 conceptual model. To describe your semantic asset you must fill in the repository, semantic asset*, semantic distribution, contact information, publisher*, licence*, documentation, and item sheets. Sheets marked with a * -symbol are mandatory. The spreadsheet also contains a number of sheets with pre-defined controlled vocabularies, such as asset type, licence type, representation technique, file format, etc. These sheets help you to select the right codes to use, but you should not update them.

2. Import your spreadsheet in Google Refine;

Install Google Refine and the RDF extension on your system following the installation guide. Create a new project in Google Refine and import the spreadsheet you created in step 1. The screenshot below indicates which parameters to fill in.

3. Add the pre-defined mappings using the Google Refine RDF extension

The Google Refine RDF extension allows you to determine the intended structure of the RDF data by drawing a template graph. You can manually edit the RDF Skeleton by clicking “RDF” > “Edit RDF Skeleton...” in Refine.

To save you the trouble of adding a mapping from the spreadsheet to the ADMS RDF vocabulary, you can re-apply the operation history. The operation history contains all operations previously carried out  to build the RDF Skeleton for the spreadsheet template. To apply the operation history:

1. Click “Undo/Redo” > “Apply...”
2. Paste the Operation History from this text file
3. Click “Perform Operations”.
4. Click “RDF” > “Edit RDF Skeleton...” and verify that the RDF Skeleton has been created from the Operation History.

4. Export to RDF using the Google Refine RDF extension

Now everything is in place to generate and ADMS RDF file, containing your semantic asset descriptions. Click “Export” > “RDF in RDF/OWL” or “RDF in Turtle”. 

You now have obtained an RDF description of your semantic asset in either Turtle or RDF-XML. You can validate your RDF file using the W3C RDF Validator. You can save your metadata on your web page and share it with the forthcoming federation of semantic asset repositories  on Joinup.